Three newsworthy scoops as we exit Summer and head into peak back-to-school and then a fast down-hill ride to Holiday 26. (15 weeks until Thanksgiving).

FLASH 1: ChatGPT Product Card Spike

Over the weekend our ACO monitoring system started going bonkers on ChatGPT data:

• Product Card visibility up 20% across many clients (% of prompts that elicit a product card)

• Card count per prompt that elicits a product card spiked from a range of 1-2 to 6-8

This has persisted through today.

Subscribe now

Examples of New Experience

Here we’re looking for ‘face serum’:

Here I am a logged in user and ChatGPT knows I am in Raleigh, NC. Notice that the top of the map is dominated by a very detailed store finder. Notice the local map now is front and center, for this shopper experience. As I flip through different product cards, the beauty retailers that have the product turn on and off in the map, so it’s more of an interactive view than you would get from the google ‘show me in stock near me’ feature of Google Shopping.

There are also a ton of product cards under neath here - not in a carousel, each with a lot more exposition about the product.

When I click on a product card the offer card has changed:

The biggest change is the hero image has been enlarged compared to what we’ve seen previously and I can flip between images:

Finally, if I drill down on a store it pulls up a ‘store card’→

Here you get store-level reviews aggregated from Yelp! and what looks a lot like a “Google places” type treatment.

We’re seeing this consistently across free and logged in accounts - what are you seeing?

Leave a comment

FLASH 3: Three New Shopping Related Feature Flags Detected

We have another set of agents that monitor the chatgpt front-end code for any interesting visible changes. As mentioned previously, ChatGPT has a framework:

1. Push new UI front-end with feature flags turned off - presumably for internal testing or tight A/B controlled audience testing (or both)

2. Over time, on the back-end they can turn these feature flags on and off when they want to roll out new features with a big bang - meaning you don’t have to wait for the new code to roll out via downloads for the app-based users (web can typically roll-out instantly).

Editorial note: Because we can’t turn these on and we’re dealing with unannounced features, this is pure speculation - educated speculation, but still 100% speculative, whereas the product-card visibility FLASH changes are live, observable and measurable.

New Feature Flag 1: shopping_sidebar_enabled

ChatGPT is using a company called statsig which is a platform for running A/B tests and gating off these types of new features. Speculating here, but we can see two statsig experiment IDs associated with this one. This pattern usually means - this feature is coming, let’s A/B it as we slowly roll it out and see which one has the highest KPIs (traffic/clicks/sales/cart adds/likes - could be anything).

New Feature Flag 2: ‘AllProducts’ Setting

The other setting is in the code in the settings/admin area. The sky’s the limit on this one. It could be a ‘god mode’ feature where super ChatGPT admins can see all the products in the system, it could be a user setting where you can see a wishlist type functionality, it could be a ‘product history’ view. This one’s harder to predict.

New Feature Flag 3: 'navigate_checkout’ and ‘prepare_checkout’

These are part of the checkout sequence. The current checkout situation at ChatGPT is basically what we call cart injection (where available down-stream). Most likely this does not indicate that they are revisiting a full-blown checkout, but is some process improvements to the injection. That being said, we haven’t seen this much action since March 26 when checkout was abandoned.

Subscribe now

META’s Crawler goes Crazy - What are they building?

Popular ‘build in the open’ account on X, levelsio on August 6 posted that he was seeing across his various sites. He then followed up with this post:

Levelsio claims that META told him they are building their own search index. EU GEO company Peec chimed in:

Cloudflare Radar tracks bot traffic across their massive network and they conveniently break it down by 'category of target’ and they have “Shopping & General Merchandise” in the 7 day window, it’s clear that META is crawling the retail category especially heavily.

Zooming out and looking at 4 weeks, you can see another interesting trend:

Look at that Claudebot activity. Why do you think Claude AND META are both showing a dramatic increase in traffic right now?

I have some guesses on what’s going on, I suspect in the next 2-3 months there’s going to be some pretty exciting announcements along the lines of my original 2026 predictions.

On the META side, digiday reports that prior to 1H 2026, the categories were balanced, but now S+GM is significantly the majority of the crawls, showing this was not ‘business as usual’.

Summary

To Summarize it’s mid August and:

• ChatGPT is adding local inventory to their product and offer cards and showing way more cards than previously

• They are in test with a shopping_sidebar feature on the consumer side.

• There’s a merchant or internal feature called “AllProducts”

• Two new checkout variables or tests were introduced

• META and Anthropic are crawling shopping sites at a frenetic pace

Stay tuned, we’re tracking this closely…

Subscribe now

Subscribe now

Editorial note: As a reminder, we prefer news items that surface new data and insights. We look for both observed data (usually in the form of permissioned panel data) and stated preference data (usually in the form of survey data). We believe no one set of data is perfect, but by looking at a landscape of datapoints, together we can spot important directional trends and shifts in consumer behavior before they happen. We’ll always stated which of the above the article incorporates at the top of the coverage.

Mastercard spent its entire earnings call on agentic commerce

Mastercard Q2 2026 earnings call · July 30

I listen to a lot of earnings calls and have to admit that prior to this quarter, I had never listened to a Mastercard earnings call. The day after the call, my Agentic Commerce alert system went through the roof about this call so I went back and listened and was blown away about not only the density (about 80% of the call by my guess) but the enthusiasm from the tone and body language of the call around Agentic Commerce. I’ve done these calls before, these are the most forensicly reviewied and lawyered content you’ll ever hear. I could fill 5 pages with the quotes from this one, here’s the top of the top: (These are all Mastercard’s CEO Michael Miebach - bolding is mine).

1. "Agentic Commerce is the next evolution in payments where the importance of security, transparency and control only increase. Agentic Commerce creates a significant opportunity for Mastercard. It leads to incremental transactions and even more opportunity for our services."

2. "It's unique because we have additional capabilities that we put into the agent pay protocol from us, and one of them is verifiable intent, which allows you to basically challenge a transaction, say, I never wanted to buy this and then the chargeback process can kick back in. This was innovated together with Google."

3. "The rise of Agentic Commerce also brings about an entirely new class of payment use cases, machine-to-machine payments. This is an expansion of our addressable market and one that we are at the forefront."

Mastercard has built the chargeback path for agent purchases. You don’t build the exception path for something you think is three years out.

Their strategic framing is a virtuous cycle: more switch transactions, more data, more value-added services to attach - all extended to agent-initiated payments. And they’re explicit that they think cards prevail in an agentic world (I tend to agree, consumer behavior is increasingly stubborn to change.

→ Mastercard Q2 2026 earnings call transcript

→ Agent Pay for Machines launch release

Keep an eye on Mastercard, they are up to big things in this space and they went from more of a defensive stance with Wall St. (“We’re going to be ok guys!”) to pure offense: (“This blows up our TAM, we’re in front, we’re adding new capabilities, we’ve partnered with Google, LETS GO GUYS!”) Note: These are my interpretations not their quotes.

Visa Declares Agentic Commerce a Growth Driver

h/t to Marcel Van Oost for putting out these sweet earnings charts.

Visa also decided Q2 2026 was the time to take a similar stance.

Two networks, one quarter, on the record. Whatever you thought about the timeline before, the rails have stopped debating it.

The contrast that makes it land came three days earlier → PayPal’s new CEO published five strategic priorities and agentic commerce wasn’t one of them.

We’re seeing a big divergence from the card companies and PayPal, you can easily guess which path I think is the right one.

→ Visa declares agentic commerce a growth driver

Similarweb Clarifies What Happens After AI Recommends You…

Similarweb — “The Downstream Impact of AI Visibility” · release July 12 with accompanying webinar July 16

This is observed data from Similarweb’s massive panel.

The key findings:

• AI-recommended brands got 2.5x more visits within seven days.

• AI-influenced visitors averaged 12.0 pageviews and 11.8 minutes on site, against 6.5 pages and 5.6 minutes for everyone else.

• 55.9% of that downstream traffic arrived as branded search. Only 8.8% arrived as a direct AI referral link.

That last one is the story. Roughly nine out of ten visits an AI recommendation generates land in your analytics wearing someone else’s clothes - branded search gets credited to SEO, direct gets credited to nothing. Anddd.. we’re back to the age-old attribution problem.

This all but verifies the Dark Agentic Commerce Traffic we’ve been talking about. Similarweb frames it as users coming back later. I’m sure in the panel data it looks this way, but it is not.

The companion webinar re-iterated another number we’ve seen over time: zero-click is now 65–68% of Google searches and it’s heading to 100% rapidly.

→ Similarweb — The Downstream Impact of AI Visibility

McKinsey and ICSC: US Agentic Retail Forecast: $1 Trillion

McKinsey & ICSC - “Shopping in the Age of AI: Redefining Stores for a New Era” · webinar July 13 and Shopping in the age of AI paper

McKinsey uploaded a YouTube and it’s a good watch. this data is based on stated preference from 3,004 US consumers plus C-suite interviews and their global expert network, across grocery, fashion, health and beauty, home, electronics, and food and beverage.

The headline number is up to $1 trillion in US B2C retail revenue through agentic commerce by 2030. But the stat that actually changes how you think about urgency is this one → the top decile of retailers is projected to capture more than 85% of sector economic profit.

McKinsey believes: “agentic commerce concentrates.”

The Mckinsey store paper adds:

Their convenience-versus-discovery framework is a genuinely good model - physical retail bifurcating rather than declining, with a single store unable to optimize for both missions at once. And they name the other macro forces stacked behind it: ~$100T of intergenerational wealth moving to Gen Z and millennials, and ~40% of US workers still hybrid or remote, which re-anchors shopping patterns near home instead of the office.

This bit hits home: “agents compare your store to competitors before the consumer ever leaves home, and data accuracy is table stakes”. Not only does your online catalog need to be upgraded, but you need to make sure agents looking at local inventory have access to as real-time store-level inventory as possible.

→ McKinsey — Shopping in the Age of AI

LDWW Survey: 70% of Americans are already shopping with AI

LDWW survey of 2,000 US adults - stated preference

Here’s what this survey reveals:

• 70% already use AI to help them shop, and plan to use it more

• 71% expect to increase their use in the years ahead

• 67% say AI influenced a recent purchase decision

• 30% have already bought products or services through AI, not just researched

• 19% say AI tools are the most trustworthy source they use to shop

Nearly one in five Americans trusts a language model more than any other shopping input they have.

→ LDWW — 7 in 10 Americans Now Use AI to Help Them Shop

Similarweb’s 2026 Gen AI Landscape

Similarweb — “Gen AI Landscape” (observed panel data)

Similarweb used to share the highlights of their panel data around GenAI monthly then stopped about 6 months ago. This is the first GenAI mega update since, and the best single map of where AI search traffic is actually going.

The two charts are the tip of the iceberg, here’s the Alexa shopping data for example:

→ Similarweb — 2026 Generative AI Landscape

Gener8’s AI Landscape

Speaking of panel data, Gener8 sits on a big panel dataset and the July analysis has some genuinely interesting trends from the data:

(Note: I’m not sure what they mean by AI Mode being the default, that is not the case at least in the US), but they are making it increasingly easy to hit the ai mode button.

Gener8 Ai Landscape July 2026

7.92MB ∙ PDF file

Download

Download

That’s the End to July 2026

As we head to the end of Summer and beyond not only is the information’s going to be coming out fast and furious, but the big answers engines have been quietly building and I predict we’re going to have a bunch of news after Labor Day so buckle up, it’s going to get busy.

Subscribe now

Subscribe now

Colin Sebastian is Baird’s Managing Director of Internet Research, one of the leading sell-side analysts covering e-commerce, social media, and online advertising (and video games!). His coverage universe runs from the mega caps (Amazon, Alphabet, Meta) through mid caps (Airbnb, Shopify, eBay, Instacart) down to smaller names like Pattern.

He started in investment banking, “saw the light” and moved to equity research, and has now covered internet at Baird for over 15 years.

Listen/Watch Our Interview:

This week Scot goes on a markets deep dive with one of Wall Street's leading internet analysts. It’s a tour through the mega-cap playbook (Meta, Amazon, Google, eBay, Shopify) and how professional investors are actually pricing the agentic-commerce shift. Colin covers over $12 trillion of market cap across five names alone, so this is the view from the models, the channel checks, and the trading desk.

This is a wide-ranging, numbers-forward conversation about where the money moves as agents become the buyers and which companies Wall Street thinks come out ahead.

Highlights

• 𝗧𝗵𝗲 𝗦𝗣𝗘𝗘𝗗 𝗳𝗿𝗮𝗺𝗲𝘄𝗼𝗿𝗸: Baird’s formula for internet winners: Scale, Product, Engineering, Efficiency, and Data. Nail all five, and the competitive advantage compounds.

• “𝗟𝗲𝘁 𝘁𝗵𝗲 𝗮𝗴𝗲𝗻𝘁 𝗱𝗼 𝘁𝗵𝗲 𝘄𝗮𝗹𝗸𝗶𝗻𝗴”: Colin explains why skeptics are dramatically underestimating the size of the opportunity.

• 𝗠𝗲𝘁𝗮 𝗶𝘀 𝘁𝗵𝗶𝘀 𝘆𝗲𝗮𝗿’𝘀 𝘁𝗼𝗽 𝗽𝗶𝗰𝗸: Colin’s highest-conviction AI bet is Meta driven by WhatsApp business agents, Meta AI, and unmatched consumer reach. (Last year’s pick: Google.)

• 𝗠𝗼𝗱𝗲𝗹𝘀 𝘄𝗶𝗹𝗹 𝗯𝗲𝗰𝗼𝗺𝗲 𝗰𝗼𝗺𝗺𝗼𝗱𝗶𝘁𝘆: The real value won’t live in foundation models, it accrues to the platforms, applications, and distribution layers built on top.

• 𝗔𝗺𝗮𝘇𝗼𝗻’𝘀 𝗼𝗻𝗲-𝘄𝗮𝘆 𝗱𝗼𝗼𝗿: By opening its catalog to AI agents, Amazon may have crossed a line it can’t reverse and that could strengthen its ecosystem even further.

• “𝗖𝗹𝗼𝘀𝗶𝗻𝗴 𝘆𝗼𝘂𝗿 𝗲𝘆𝗲𝘀 𝘁𝗼 𝘁𝗵𝗲 𝗯𝗶𝗹𝗹𝗯𝗼𝗮𝗿𝗱𝘀”: If AI agents skip search results and category pages, they also skip the ads, putting pressure on retail media’s biggest revenue stream.

• 𝗪𝗵𝗼 𝗼𝘄𝗻𝘀 𝘁𝗵𝗲 𝗮𝘂𝘁𝗵𝗲𝗻𝘁𝗶𝗰𝗮𝘁𝗲𝗱 𝗰𝗼𝗻𝘀𝘂𝗺𝗲𝗿 𝘄𝗶𝗻𝘀: The long-term winners will control identity, payments, checkout, and fulfillment, not just discovery.

• 𝗪𝗵𝘆 𝗪𝗮𝗹𝗹 𝗦𝘁𝗿𝗲𝗲𝘁 𝗶𝘀 𝗻𝗲𝗿𝘃𝗼𝘂𝘀 𝗮𝗯𝗼𝘂𝘁 𝗦𝗵𝗼𝗽𝗶𝗳𝘆: AI could bypass storefronts and even Shop Pay. That uncertainty alone is enough for investors to discount the stock.

• “𝗪𝗲’𝗹𝗹 𝘀𝗲𝗲 𝗶𝘁 𝘁𝗵𝗶𝘀 𝗵𝗼𝗹𝗶𝗱𝗮𝘆”: Colin expects agentic commerce adoption to meaningfully accelerate after Q4, even a one- or two-point lift on today’s numbers would be significant.

Colin is one of the sharpest lenses out there on how the market is actually pricing all of this. Enjoy the conversation!

🎧 Listen wherever you get your podcasts: Spotify, Apple, etc.

Or, we have a video version on YouTube (make sure to subscribe!) →

Want to Learn More?

Stay on top of what’s happening in agentic commerce, from the tools powering it to the founders building with it by following us across platforms:

• Podcast: https://retailgentic.transistor.fm

• YouTube: https://www.youtube.com/@Retailgentic

• Instagram: https://www.instagram.com/retailgentic/

• X: https://x.com/retailgentic

• TikTok: https://www.tiktok.com/@retailgentic

We're here to make sense of what’s next in retail + AI and bring you along for the ride. Subscribe and stay tuned!

Most of what we cover in these digital pages, is the Micro picture - meaning very specific things in the our World of Agentic Commerce which is a huge opportunity in it’s own right. That being said, in this post, we’re going to cover both the Micro and the Macro for the first time really, so I wanted to first explain the ‘why’.

Everything we talk about here at Retailgentic around Agentic Commerce sits on top of an even bigger wave of innovation and capital, that I’ll call the AI Buildout which is supported by the Wall St. AI Trade. That’s the Macro. Let’s start at the Micro because that’s what you come here for and if you’re interested in the Macro, keep reading 😄

In this post we’re going to focus on two specific parts of Amazon’s Q2, if you want to do a complete deep dive, check out:

Subscribe now

Jason and Scot Amazon Q2 Deep Dive→

Watch here→

Listen here→

(back to the post)

The Micro: Andy Jassy’s Comments on Rufus Alexa Shopping

On the Rufus side, there were four datapoints that were interesting. In typical Amazon style, they are ‘just enough’ to get you excited, but leave you with more questions than they answer.

The big four Alexa quotes were:

• More than 350 million customers used Alexa Shopping during the past 12 months.

• In Q2, active users nearly doubled, while interactions increased more than 5× year over year.

• U.S. customers who use Alexa for Shopping spend more than 40% more per order.

• Shoppers clicking a sponsored prompt converted to a sale 48% more often and spent 21% more on average than shoppers who did not.

This is amazing and it’s clear Alexa Shopping is doing really well. I’ve learned over 30yrs to listen to Amazon’s words, but to put 80% of the focus on their actions. The fact that in the last year we have prompt pills all over the discovery and PDP experiences and most recently we have

We’re going to be on-site at Amazon in mid August and here’s the questions we’re going to be looking to figure out answers for:

• 350m customers - Is this global/US? What % of customers does this represent?

• Q2 2x active users and 5x y/y interactions - Was this 2x from Q1 or from year ago? Interactions - is that prompts, does it include clicks, etc?

• Alexa Shopping AOV up 40% - Is there selection bias in here (you take amazon’s best clients, put them in a cohort, that’s an aov that’s 40% higher) - or there’s a more scientific ‘hold back group’ A/B test going on here?

• Sponsored prompts - wow!, but again, selection bias?

As I said, I can tell there’s really something here, because of the actions, but I’d love to have the dots connected a lot more here.

Stay tuned, we’re on it. Now let’s zip back up to 30,000 ft for the Macro…

The Macro: Jassy Single-Handedly Shows Wall St the Clear Path to Huge ROIC on AI Spend

Jassy’s Legacy at Amazon

Foundational to this part of the story, is understanding just how much credibility and Jassy has with Wall St. Jassy isn’t an Amazon newbie, he started in 1997. Amazon urban legend has it that in 2003 he came to Bezos (or they co-developed the idea) for what would become AWS. This is now a huge driver of Amazon’s business. In 2016 he was officially tagged as CEO of AWS and then in July or 2021, Jassy took over the ‘main’ CEO gig.

In that period of time, Amazon spent more and more on Data centers and servers and Amazon showed them how the company has become world class at ROIC - return on invested capital. If you build a $2B data center, how much revenue does that produce over it’s life? What’s the break-even period? Jassy has 23yrs of proving out, via AWS that he is a great steward of investor’s capital and will invest it with an eye towards getting a payback as fast as possible.

Share Retailgentic: Agentic Commerce meets Retail and Brands

Setup going into the call…

The big AI companies, Meta and Hyperscalers are spending so much on AI datacenter and GPUs that it has created a huge wall of worry on Wall St. that can be summarized as: “What if all this capital being deployed is a bubble? What if it' never generates an ROIC? The market has been searching for a clear sign that there is ROIC. Heading into Q2:

• Google Announced -increased their capex and had a great, Q, but didn’t answer the questions the market has - neutral.

• META Announced - Zuck basically said - I want to spend a lot more and to make it work, maybe we’ll do, I don’t know like enterprise apps and stuff! The market absolutely hated this - and 🤮 all over the place.

• Microsoft Announced - this was better, Azure accelerated and Co-pilot seats increased an astounding 50% Q/Q to 30m. The market liked this, but still, it’s hard to tie this to AI spend - nobody connected the dots.

That’s how the table was set going into Amazon’s call, which I’m calling:

Jassy’s Shifts Into AI MAXXING Mode!

Amazon’s AWS results crushed expectations from a y/y growth, but on top of the revenue expansions/beat, margins EXPANDED. As if that wasn’t enough…

Jason and I have 37 of these Amazon quarterly recaps in the rear view window, we have only missed 2, yes 2 in the last 10 yrs. I listen to every conf call, study the transcripts, read at least 5, sometimes 10 wall st reports, read the annual letters and so on. In the Bezos and Jassy eras, I have never heard a call that had so much unbridled enthusiasm as this call. Written words take away a lot of the audio clues/’body language’, but with that back drop. Here are the biggest quotes from the call:

• Booming - “Q2 was another very strong quarter for Amazon. I’ll start with AWS, which is booming right now, and I’ll share the numbers, what we think is going on and why we’re enthusiastic about the ROIC equation, even with heavy CapEx the next few years.”

• ROIC? - “At this level of spend and higher, we have clear line of sight to strong financial returns.”

• ROIC Dots…Connected! && Breakeven…revealed! - “For servers and networking equipment, on average, it takes a little less than 3 years to break even on that investment.” He added that the equipment has a useful life of at least five to six years, while most AI capacity is being contracted for at least five-year terms—creating significant free cash flow during the two to three years following breakeven.

• AI vs. CPU correlation - “We see the margins and returns in AI tracking what we saw with core at the same point of evolution, actually a little ahead.”

• Icing on the cake - “The resulting revenue, free cash flow and return on invested capital is very compelling.”

This chain of five statements was EXACTLY what Wall St. was looking for PLUS it was delivered by Jassy who has a world of credibility PLUS he delivered it with enthusiasm PLUS he delivered it extremely clearly and coherently.

The Icing on the Icing on the Icing on the cake….

As if that wasn’t enough, Jassy put the icing on the icing and said:

Conclusion

To Summarize: Booming, ROIC, Breakeven, Accelerating revenue, growing margins and AWS could be a $1T business. Phew, that was a lot.

Today the market took a breather and there was other dramatic Google news (a topic for another day), but prior to that, Amazon was up 12% which at their ~$3T market cap is $360B (or more than 5 Target’s in market cap growth to put it in comparison).

On top of that Google, and Microsoft, the other Hyperscalers were up too, because Jassy helped show the ROIC on ALL AI investment, not just Amazon’s. I haven’t done the math, but that could be highest $/hr ever recorded in value created from that conference call.

Subscribe now

Subscribe now

We’re excited to have Shopify on the pod! Mani Fazeli is VP of Product at Shopify, leads agentic commerce, and owns a huge slice of the core commerce loop.

He’s spent a little over six years shaping the core commerce loop with merchandising, checkout, customer accounts, B2B (he brought B2B on Shopify to life), plus work on natively supported stablecoins, and for the last ~15 months has led agentic commerce.

Listen/Watch Our Interview:

This is a platform-and-stats deep dive: Scot gets Mani to pin down what agentic commerce actually is (with a four-stage framework that’s the cleanest map we’ve heard), then pulls out a run of genuinely wild numbers on what Shopify is seeing across millions of merchants, before going deep on the catalog, the protocol soup, and everything that shipped in the Spring ‘26 Edition.

Highlights

• Crawl, walk, run, fly: The framework of the episode. Crawl is AI-driven discovery (the biggest, most impactful stage today); fly is fully discretionary, budget-based buying. Mani’s verdict on where the industry actually is: “somewhere between walk and run today.”

• 8X traffic, 13X orders: Shopify’s Q1 numbers… AI-driven traffic to merchants up 8X year over year, and orders through those channels up 13X.

• AI buyers just hit harder: New buyers who discover a merchant through AI order at roughly twice the rate of organic search, are about 50% more likely to convert, and spend about 14% more per order because they show up deeper in the consideration funnel.

• 70% of AI orders are long tail: About 70–71% of AI-attributed orders last year fell outside the top 100 product categories. Models “don’t need to learn how to reach for the small merchant, they just need to learn the intent of the buyer.”

• 2X on fresh data: The Shopify catalog, over a billion products across millions of merchants, converts about 2X better than stale or scraped data, thanks to real-time pricing and availability straight from the merchant.

• A “7 out of 10 at everything”: On how AI is reshaping product, design, and engineering roles.

• The non-merchant-of-record era: Don’t become the marketplace reselling goods, just be the interface to the merchant. The merchant keeps the customer relationship and runs their own checkout.

• UCP was a “labor of love”: Shopify was at the genesis of the Universal Commerce Protocol.

• Sidekick, Pulse & Autopilot: Sidekick usage is up about 4X year over year; Pulse turns Shopify’s data into proactive recommendations right on your admin homepage; and the guiding principle stays “always put the control in the merchant’s hand.”

• B2B may reach “fly” first: By 2030, Mani splits the world into DTC and wholesale/B2B and bets B2B, where both sides transact professionally and obsess over efficiency, may hit full automation before consumers do.

Mani’s four stages are the cleanest map of agentic commerce we’ve heard, and the numbers behind them are frankly staggering. And as he reminds us, we’re still in the early innings. Enjoy the conversation!

🎧 Listen wherever you get your podcasts: Spotify, Apple, etc.

Or, we have a video version on YouTube (make sure to subscribe!) →

👉 And check out the Shopify newsroom blog here

Want to Learn More?

Stay on top of what’s happening in agentic commerce, from the tools powering it to the founders building with it by following us across platforms:

• Podcast: https://retailgentic.transistor.fm

• YouTube: https://www.youtube.com/@Retailgentic

• Instagram: https://www.instagram.com/retailgentic/

• X: https://x.com/retailgentic

• TikTok: https://www.tiktok.com/@retailgentic

We're here to make sense of what’s next in retail + AI and bring you along for the ride. Subscribe and stay tuned!

Here at Retailgentic, we just completed our 4-part ACO Everywhere series (starts here if you missed it) and fundamental to that series was the concept of context capture. Once we wrapped it up, I hopped on a plane and went out to NRF Nexus to speak and hang out with all my favorite brand and retailer people and talk about all things Agentic at the Think Tanks and in the halls.

After the first several conversations, I realized we have a big problem in the industry that spans across brands, retailers and all the vendors…

Subscribe now

The Problem

Ask ten vendors, ten brands and ten retailers what “product enrichment” means.

You’ll hear:

• Attribute completion

• PIM automation

• Image tagging

• Catalog distribution/endpoint publication/MCP-ification

• AI-generated descriptions

• Taxonomy mapping

• That thing my PIM does?

They’re all called Enrichment, yet they are all solving different problems and actually taking very different actions around your product catalog.

What I discovered is that like “Agentic Commerce” itself, we have overloaded one phrase (enrichment!) to describe five completely different capabilities.

The Enrichment Gap: Introducing the 5 Levels of Product Enrichment

Instead of thinking about enrichment as “better data,” think about it as increasing levels of shopper understanding coming from the marriage of shopper intent and product data.

Level 1 — Structured Basics

This is what most vendors do and where merchants, especially brands, are investing today:

Examples:

• Dimensions

• Weight

• Color

• Material

• Brand

• UPC

• Images

These basic attributes are obviously necessary table stales, but add zero differentiation.

Level 2 — Conversational Attributes

In Level 2 - it’s starting to get interesting because in giving robotic: “material=cotton” …now we’re answering questions.

We’re answering:

• Is this breathable?

• Will this shrink?

• Does it wrinkle?

• Is it comfy?

• Can I wear it in the summer?

• Is it good for travel?

The shift from Level 1 to Level 1 is we’re now enriching conversations, not just product data.

It’s important to note that this is just not me ‘guessing’ or making a prediction - Google is focused on getting this data from merchants for getting brands and retailers to this level and it’s part of UCP.

Level 3 — Shopper Context

Because the Answer Engines have tons of shopper intent, on top of basic product data and conversational attributes, they are looking for product-level shopper context from authoritative sources (like the brand and their retail partners).

Shopper context like:

• Use cases

• Occasions

• Compatibility

• Seasonality

• Product lifecycle

• Pros

• Cons

• Selling points

• Related products

At Level 3, we’re past attributes and into buyer context.

Level 4 — Market Context

Products don’t live in isolated, they exist in markets, markets that change a lot and that rate of change is accelerating.

Market Context includes:

• Promotions

• Popularity

• Inventory

• Trends

• Reviews

• Price changes

AI agents care enormously about these and it’s a top input into who wins the offer card.

Level 5 — Agentic Context (new)

At the final, most advanced level of enrichment, it isn’t about products, it’s about decisions.

Examples:

• Which product best fits this shopper

• Which product best fits this prompt

• Why should the AI recommend Product A over Product B?

• What objections must be overcome?

• What evidence supports those claims?

• Which customer segments respond best?

Now you’re optimizing for reasoning instead of 2002 basic database filtered navigation.

Why This Matters

In these digital pages we’ve covered these first principles of agentic commerce:

• Consumer behavior has changed more in the last 3 years than 20 years of keyword jail, causing us to over-optimize for keywords

• Agentic commerce optimizes to matching the large amount of shopper intent with the product context.

• Multi-surface context capture is your competitive advantage and your moat.

• All of this has to be done in a closed loop, or you aren’t optimizing, you’re guessing.

The New Definition of Enrichment: Agentic Enrichment

Product enrichment is no longer the process of plugging in missing attributes.

In the Agentic Commerce Era, it is now the process of continuously adding decision-ready context that enables AI to confidently recommend, compare, explain, and sell the product.

How to tell if you aren’t enriching for Agentic Commerce…

• Each product doesn’t have a detailed and growing FAQ (substantially longer than what’s on the PDP)

• Your vendor/team doesn’t know that UCP/ACP also have a product catalog spec in addition to the payments piece

• You’re using an off-the-shelf AI model to recommend and fill in data (it will not make good recommendations and it will fill your catalog with reddit-fueled AI slop)

• They aren’t talking to you about how to get access to more context

• You aren’t enriching in your company’s ‘voice of brand’

• You’re sending the same data without channel customization to Google, ChatGPT, CoPilot and Meta.

Conclusion

The companies that win Agentic Commerce won’t necessarily have the biggest catalogs or the happiest PIM that has every mandatory field filled out to perfection. They will have the richest context that is continuously optimizing and responding not only to underlying product changes, but shopper behavior and market dynamics fueled by context capture across as many surfaces as possible.

THAT’s the enrichment gap.

Subscribe now

Subscribe now

Kartik Hosanagar is the John C. Hower Professor of Technology and Digital Business at the Wharton School, where he also runs the university’s AI center. He’s an academic and entrepreneur of roughly 25 years: he co-founded and built the core IP for Yodle, a venture-backed local online-marketing company later acquired by Web.com, and was an advisor to Milo (local inventory, acquired by eBay).

Kartik’s the author of “A Human’s Guide to Machine Intelligence” and an angel investor in a few dozen startups. Most recently he co-founded Bodhium Labs, an applied AI lab whose first product runs “counterfactual simulations” to help brands understand and influence how AI engines perceive and recommend them.

He also has a Substack that goes deeper into AI and business. Check it out 👇

Creative Intelligence | Kartik Hosanagar

The intersection of AI, creativity, mindfulness, and entrepreneurship

Listen/Watch Our Interview:

Kartik Hosanagar wears a lot of hats: Wharton professor, serial entrepreneur, angel investor, and now co-founder of an applied AI lab building simulation tools for AI search. Scot and Kartik land in almost exactly the same place on where agentic commerce is heading, but they get there from opposite directions: Scot on gut and operator instinct, Kartik on research. The result is a wide-ranging conversation about treating AI as a customer, the levels of agentic commerce, and what brands should actually be doing this week, this quarter, and this year.

Highlights

• AI is a new customer category: Once shopping goes agentic, Kartik argues you have to ask of AI what you ask of any customer: how do we inform it, persuade it, and what biases does it have?

• The driverless-car model: Agentic commerce has levels 0–5, just like autonomous vehicles. Level 5 is “buy the shoe and make sure it arrives in a week” hands-off; Instacart-style automation and AI-enabled discovery are lower levels and we’re already in the era.

• The funnel got outsourced: Top-of-funnel work customers used to do themselves like discovering brands and building a consideration set is now done by AI, so shoppers arrive far later in the funnel. Discovery is “fundamentally transformed.”

• 4 words → 60 words: From his CommerceNext keynote, average search went from about four words on Google to about 60 words on the first ChatGPT turn, then multi-turn. Nobody types “men’s running shoes” anymore; shoes show up on conversation turn seven.

• UCP, ACP and the protocol wars: Kartik explains Google’s UCP (product feeds today, transactions later), why OpenAI backed off ACP instant checkout, and reads Amazon joining UCP as “testing the waters,” not a strategy reversal

• Counterfactual simulations: Bodhium builds “digital twins” of ChatGPT and Gemini and spins up virtual worlds where your site looks different, testing a thousand hypotheses in minutes to find what makes AI answers more accurate and favorable. Market research, but for AI.

• The today / this quarter / this year roadmap: Now: audit conversations, visibility, and which bots read your site. This quarter: build simulators and a UCP/MCP workstream. This year: scale it with agentic workflows.

• Two front doors: Kartik doesn’t think websites disappear, but “a website is not the right front door for an agent.” Expect a human web door and an MCP door for agents, plus omnichannel behavior where each feeds the other.

Kartik calls himself an "agentic commerce maximalist," and after this conversation you'll see why... the research and the gut instinct point the same direction. Enjoy the conversation.

🎧 Listen wherever you get your podcasts: Spotify, Apple, etc.

Or, we have a video version on YouTube (make sure to subscribe!) →

Want to Learn More?

Stay on top of what’s happening in agentic commerce, from the tools powering it to the founders building with it by following us across platforms:

• Podcast: https://retailgentic.transistor.fm

• YouTube: https://www.youtube.com/@Retailgentic

• Instagram: https://www.instagram.com/retailgentic/

• X: https://x.com/retailgentic

• TikTok: https://www.tiktok.com/@retailgentic

We're here to make sense of what’s next in retail + AI and bring you along for the ride. Subscribe and stay tuned!

Series Recap

• Part I introduced the importance of product-level context capture.

• Part I also highlighted that Answer Engines now give merchants 4 high-capacity lanes of product-level data (via UCP/ACP standards) to input into the Answer Engines a significantly larger amount of product-data than we ever contemplated before 2026.

• In Part IIA we theorized the best places to capture context.

• Part IIB showed that in practice

• Finally, we looked at how important recursive improvement loops are in optimizing systems with AI-fueled agents.

These are building blocks, let’s start putting them together to build the next level of Agentic Commerce Optimization.

Subscribe now

Putting it All Together…

In Part IIB, we showed in detail, that for one SKU (the Columbia Watertight II jacket) if we looked at these eight surfaces for context:

1. Answer Engine - ChatGPT

2. Answer Engine - Gemini

3. Retailer Agent - Alexa

4. Website - On-site search

5. Website - Reviews

6. Website - FAQs/Q+A

7. Social Media - TikTok, Pinterest, Instagram, YouTube

8. Offline: Store and Manufacturer

In these 8 contextual sources, we found about fifty pieces of important SKU-level context:

And THAT was just illustrative. If you could theoretically build a massive ‘context capturing machine’ you’d want it to cover ALL the answer engines, all the retailer agents, the closed and open data from the website, all the social media channels, and the gigabytes/terabytes of data stuck in the offline operations.

Context Capture Machine

You can imagine it would look something like this:

Capturing valuable product-level context from every available surface and looping through it. That’s what we illustrated in IIB.

Then What?

But there’s something missing. Let’s get tactical again.

Let’s pick one of the negatives surfaced by answer engines and retailer shopper agents for our hero SKU (Columbia Jacket) to ‘fix’ - let’s work on the combined issue of zipper reliability, pocket waterproof-ness and the phone-in-pocket safety bundled together.

For this one, because all the LLM based systems have now given us the ability to send a very large FAQ, we introduce a new FAQ: (I’m going to make educated guesses on this and some assumptions to illustrate how this works. Note this is never AI Slop and never bad intent/grey hat/black hat -the brand/retailer is just stating product facts):

*****BEGIN FAQ ADDITIONS TO UCP/ACP DATA FEED AND ON PDP METADATA

Question: Are the pockets zippered?

Answer: Yes, both pockets are zippered.

Question: How reliable are the zippers?

Answer: We use an independent garment testing lab (<link to report>) and they report the zippers on this jacket are good up to 1200 zip cycles. With daily wear this would be 2yrs with several daily zips a day. Most customers wear this seasonally or in inclement weather which would extend the life of the garment well beyond 2 years.

Question: Are the pockets water proof for my phone?

Answer: With the zippers closed and thanks to our Omni-Tech fabric technology, the pockets are are very water resistent. That being said if the jacket is held under water, zippers will leak. Its also helpful to note that many modern phones are increasingly water resistent and water proof - check your model for details.

*****END FAQ ADDITIONS TO UCP/ACP DATA FEED AND ON PDP METADATA

To be clear, this does not go on the human-readable part of the PDP - you can put it there if you want

Next, we publish this FAQ to all the selling surfaces.

Finally, after a week of publishing the new FAQ version, we re-run our context capture experiment again and look to see if the zipper context changed and what new context bubbled to the top.

Note: after about a week it will change in the answer engine data, after six months the underlying models will all be trained in the new data.

By implementing the:

Context Capture→Update Product Catalog→Publish

three steps, we now have created a feedback loop that can iterate. We call this strategy ACO Everywhere:

Introducing ACO Everywhere

It’s been about 1 ‘AI Year’ (the equivalent to 5 pre-ChatGPT years) since we coined the phrase ACO and introduced our popular 9-step framework (here) . We’ve been deep in this at ReFiBuy systemizing ACO and have now optimized millions of SKUs. As we’ve learned more about the nuances of ACO, the answer engines have matured their Agentic Commerce capabilities and our customers have worked with us to experiment on the components of this concept.

This is the future of Agentic Commerce optimization for brands and retailers:

Highlights of ACO Everywhere:

• Retailers and brands can start with whatever Agentic Surfaces are most important to them.

  • Most retailers start with Answer Engines (ChatGPT and Gemini) and then add the retailer.com

  • Most brands start with the Retailer agents and then work on brand.com

• Different verticals will have different agentic surfaces and contextual captures that they will want to prioritize. For example, fashion and beauty will lean in on social context more than say, auto parts.

• You can easily manually research a handful of SKUs (we’ve given you a roadmap to do in part IIB and) then manually test this yourself and you will see the results over < 30 days. Unlike AEO/GEO/SEO, ACO Everywhere is: visible, understandable, actionable and the results are measurable.

Why Implement ACO Everywhere?

I was on a customer call recently and our VP of Research, Andrew Bell was walking a brand through the gaps in their PDP/ASIN and they asked if doing all this work would be worth it. His answer made a huge impression on them (and me):

That’s the ‘why’ - if you don’t implement ACO Everywhere, your competitor will. It’s right there for the taking. The context is ‘free’ - its everywhere and historically we didn’t have anything to ‘do’ with it, and now we do. Look back at post IIB - you can see that every time the Columbia jacket doesn’t answer a question, a competitive product enters the conversation.

Because the loops are weekly you can see results quickly and there’s 18 weeks until Thanksgiving, so there’s plenty of time to get roing - that’s 18 weekly loops!

The 11 Steps of ACO Everywhere

To wrap up the post and the entire series, we are updating our “9 Steps to Agentic Commerce Optimization“ popular diagram we introduced in Q3 2025:

We now have the 11 steps to ACO Everywhere and introduce 2 new steps:

• 10. Multi Surface Content Capture

• 11. Recursive Catalog Loops

Time to get going with YOUR ACO Everywhere Strategy:

Context Capture→Update Product Catalog→Publish

Want to Learn More?

Today at ReFiBuy we are launching a complete initiative around ACO Everywhere and have some openings for our design partner program. Details at:

ACOEverywhere.com

Subscribe now

Subscribe now

We have a Special Summer Treat for you today! Bobby Stephens and Apurva Pangam from Deloitte Digital are on the pod to bring the Retailgentic audience an exclusive first look at Deloitte Digital’s 2026 B2C Commerce Research report called: “Bridging the divide” t’s a double-sided study pairing 500+ merchant leaders with 1,000+ consumers, with two years of trend lines behind it, and the throughline is a gap: brands think they’re doing far better than their customers do. Bobby and AP walk through where consumers and merchants align, where they badly diverge, and what the leaders are doing that everyone else isn’t. You’re getting it first.

Highlights of the Report

To me the punch line to the whole thing - this matters - if merchants lean into ACO, you are rewarded financially with 60% faster growth, and 2X revenue beats according to the Deloitte data→

Where should you focus?

The above ‘perception gap’ graphic really packs a punch and it’s got a ton of data to peel the onion on. One great thing about the underlying survey data is it has been running annually for 3+ years so can track not only the current perception gap, but changes over time -that’s the black box.

The gaps from largest to lowest between merchant perception (the left bar) and consumer perception (the right bar):

• 19% delta - Accurate Site Search - My take here is Answer Engines have elevated the search experience to get answers faster with less friction. Brands are not keeping up, and in fact falling behind.

• 17% - Helpful Customer Service - Again, for answer engines, ‘time to resolution’ is dropping and there’s less friction - memory and features

• 16% - Clear Availability of Inventory - Nothing’s more frustrating than getting excited about an item to later find it’s out of stock.

• 15% - Proactive delivery updates - When am I getting my new thing?!

• 13% - Clear pricing - On this part of the site you said $X and now it’s $Y (where Y is always » X - wth?!

That’s one chart, but it really is quite interesting to see see both sides of the coin.

To help us unpack all the report, Bobby and Apurva walk us through the data in greater details. Let’s go into their background now→

About Bobby and Apurva

Bobby is Head of US Retail & CPG at Deloitte Digital His path is refreshingly non-linear: he started on the retail buying side doing men’s fashion in a department store, got tapped as “the young guy” to help launch the company’s website, and moved onto the technology side designing and building early e-commerce sites. Around 2012 he left Deloitte, raised venture money, and ran a footwear startup in Chicago for about four years before selling it, taking a year off, and returning to Deloitte in ~2016. He’s led the US retail and CPG practice ever since.

Apurva “AP” is Commerce Practice Lead & Principal at Deloitte Digital. He’s been at Deloitte roughly 20 years, focused the whole way through on customer experience, marketing, and commerce, working both B2C and B2B clients and driving large-scale digital transformations globally.

Listen/Watch Our Interview:

This is a report deep dive, and Deloitte gave Retailgentic the exclusive first look at brand-new research: hundreds of senior merchant leaders and a thousand-plus recent online shoppers, both sides of the counter, with trend data from two years back layered in. The result is a map of where brands and consumers see the same picture and where they see completely different ones. And the data is fresh, fielded this spring, analyzed over the last month, unveiled here.

They describe exactly what those leaders do differently, and why it pays off. You want the edge? This is it.

Highlights

• The basics are the new battleground: Accurate search, real inventory visibility, clear delivery dates, easy returns are no longer differentiators, just expected. And a majority of consumers say brands haven’t improved on them in two years.

• The perception gap: Only 16–18% of consumers rate brands “excellent,” while brands rate themselves in the low-to-mid 30s, a false-confidence gap that’s driving under-investment where it matters.

• 57 definitions of personalization: the “golden term” has “probably 57 different definitions”, but ~70% of consumers say personalized experiences genuinely matter. Get it wrong and you hurt trust more than you help.

• 14–15 loyalty programs: 99% of surveyed brands offer one and 80% of customers belong to more than one, the average U.S. consumer is a member of 14–15 but active in only four or five. “You are fighting for head space.”

• The cashback gap: Over half of consumers say cashback is the loyalty benefit they’d value most but only 14% of brands offer it. Customers want financial value and simplicity; brands keep adding tiers and complexity.

• The leaders’ payoff: Consumers spend 34% more with the brands that nail the basics + personalization + sticky loyalty, 57% will try a new product from them, and 55% will recommend them to family and friends.

• Efficiency over experience: The two top-reported benefits merchants cite from AI aren’t customer-facing at all, they’re better decision-making and greater efficiency. Brands are getting their own house in order first.

• Machine-readable product data: Product content is still “teeny tiny” and over-optimized for the keyword era, even as LLMs collapse the old 20-tabs-and-a-spreadsheet decision into one answer. If the data isn’t there, the agent won’t infer it, it just gives what’s available.

• Connecting the dots: Data, loyalty, and AI in isolation aren’t enough; the winners execute across all three and make deliberate choices about the purpose of the store vs. the site vs. the app vs. the LLM.

Fresh data, both sides of the counter, and a clear-eyed read on the gap between what brands think they’re delivering and what shoppers actually feel. Bobby and AP brought the goods. Enjoy the conversation! 👇

🎧 Listen wherever you get your podcasts: Spotify, Apple, etc.

Or, we have a video version on YouTube (make sure to subscribe!) →

Check out the report! → https://www.deloittedigital.com/us/en/insights/research/b2c-commerce-experiences.html

Want to Learn More?

Stay on top of what’s happening in agentic commerce, from the tools powering it to the founders building with it by following us across platforms:

• Podcast: https://retailgentic.transistor.fm

• YouTube: https://www.youtube.com/@Retailgentic

• Instagram: https://www.instagram.com/retailgentic/

• X: https://x.com/retailgentic

• TikTok: https://www.tiktok.com/@retailgentic

We're here to make sense of what’s next in retail + AI and bring you along for the ride. Subscribe and stay tuned!

We had a slight delay on Part III of the current series, so it’s going to come out next week on July 22nd. In the interim, we have a special guest post for you today! One of the best ways to understand where retail is headed is to listen to the shoppers who are already living in the future.

My daughter, Rory, is a college sophomore and a member of the first generation to grow up with a connected device in one hand and an endless digital shelf in the other. When she sent me this guest post, one idea jumped off the page:

Social creates the spark. AI handles the questions. Commerce gets compressed.

That is agentic commerce from the shopper’s point of view - not a protocol or a fancy diagram, but a conversation that feels completely natural. Rory sees the shift because she and her friends are living it today.

Subscribe now

Back to school: I don’t know just ask chat

Retailgentic guest post, By Rory Wingo

I’m Rory, and I’m a sophomore at the University of South Carolina (Go Gamecocks!). One of the perks my university offers is a free ChatGPT subscription for every student. What started as a tool to help us study has quietly become something much bigger.

My friends and I don’t even say the formal name “ChatGPT” anymore. We just call it Chat. We customize Chat with different voices, personalities, and favorite emojis. Then we spam it with every question known to man.

• Need help updating your skincare routine after dorm life with hard water? IDK, ask Chat.

• Trying to find makeup that survives the scorching South Carolina heat on game day? Ask Chat.

• Decorating a tiny apartment on a college budget? Ask Chat.

• Trying to figure out what to wear for sorority recruitment based on a Lookbook? Ask Chat.

• Need a second opinion before buying something? Rely on Chat.

After a whole year of relying on Chat to help me curate my life, I looked around and realized my peers were using AI the same way. My age group is special: we’re the first to grow up with iPads in our hands and the internet open while making our Christmas lists instead of circling catalogs.

Shopping is changing, and it’s starting with savvy Gen Zers.

Shopping Has Always Been (and will always be) a Conversation

Growing up, shopping was never something I did alone. At the mall, I’d ask my mom which shirt looked better. Shopping with friends meant trying on clothes and eagerly waiting outside the fitting room for everyone’s opinion. Even online, I’d send screenshots to group chats asking, “Pick one” or “Is it worth $30?”

Shopping has always been a social activity. What’s changed isn’t that college students suddenly trust AI more than people, it’s that AI has become another more reliable person to ask.

When I ask Chat if a dress fits my style, which makeup will last through a football game, or what furniture will actually fit inside my apartment, I’m doing the same thing I’ve always done. I’m asking for a second opinion.

The difference is Chat remembers every conversation we’ve had. It already knows I like neutral colors. It knows my jewelry is normally gold. It knows I’ll spend a little more if something is going to last. It remembers projects I’m working on and purchases I’ve already made.

Unlike my friends, I don’t have to explain all of that every time I ask a question, Chat knows the context and history.

That’s why Chat became my first stop before Google.

Google can give me information < Chat gives me advice.

My AI Knows How I Shop Better Than I Do

While writing this article, I asked Chat to describe my shopping habits. I expected generic advice. Instead, it described patterns I had never noticed.

It pointed out that I almost always begin shopping by defining the constraints before looking at products:

• “$500 budget.”

• “Small apartment.”

• “Needs to fit a full-size bed.”

• “I need this to last.”

• “This has to help me with work.”

Then it explained something else I hadn’t realized: I don’t buy impulsively. I compare and ask questions. Before clicking Buy, I usually ask:

• Is this actually worth the price?

• What are the other options?

• What are the cons?

Chat also noticed that I don’t automatically buy the cheapest option. If spending a little more means I’ll use something for years, I’ll usually do it. Most of my purchases support something already happening in my life like my apartment, gardening, or even needlepoint.

Finally, it pointed out that I evaluate purchases almost like investments:

• Will this save me time?

• Will it improve my room?

• Will it help my career?

• Will I still use it three years from now?

If you had asked me ten minutes before this exercise how I shopped, I never would have described myself this way. But Chat recognized the patterns because it had seen hundreds of conversations over time. Whenever I ask for product recommendations, it automatically organizes the answer around this existing decision pattern.

That was the moment I realized AI wasn’t just recommending products. It understood how I make decisions.

Subscribe now

The Social Platforms Are Changing Too

This isn’t happening only because students are prioritizing AI. The platforms themselves are encouraging it.

Recently, while scrolling Instagram, I was served an ad from ChatGPT promoting its image capabilities through a personal color analysis. It wasn’t advertising a specific clothing brand or a makeup company. It was advertising the concept of asking AI which colors look best on you before you even start shopping. This really stands out to me.

Instagram ad: “Create your personal color analysis with ChatGPT Images.”

For years, social media companies competed to show us products with ad placements. Now they’re also signaling that we should consult AI before making a purchase.

Instead of simply showing me another outfit, Instagram was suggesting I let AI help decide which outfits would actually work for me.

The recommendation itself is becoming part of the product.

A Viral TikTok Lip Combo Made Me Realize What’s Changing

The biggest example happened while doomscrolling.

I saw a TikTok featuring a lip combo that everyone seemed to be recreating.

The TikTok that started the shopping journey.

Before Chat, I probably would have: paused the TikTok→ searched through hundreds of comments→ opened Google → checked Sephora → searched TikTok Shop → compared prices, watched more reviews, and eventually decided where to buy everything.

Instead, I clicked the little circle that opens TikTok’s AI assistant and asked:

“What lip combo is this?”

Within seconds, it identified every product in the video, with the right shades.

TikTok’s AI assistant identifies the products and shades in the video.

Then I typed:

“Where can I buy them?”

Instead of making me search multiple websites, it showed me retailers, pricing, and availability.

The assistant moves from identification to retailers, pricing, and availability.

here’s the second page→

Next I asked:

“Is there a TikTok Shop option?”

Rather than simply answering yes or no, it surfaced TikTok Shop listings, recommended exclusive bundles, displayed customer ratings, suggested tutorial videos, and even prompted me with additional questions I might want to ask next.

From the original video to shoppable products and bundles - in one conversation.

What struck me wasn’t simply that AI answered my questions. It got to the product discovery in seconds.

It was that it naturally handled the same conversation I would have had with a friend while we were getting ready. Offering me where to get and how they personally use the product.

Subscribe now

AI Isn’t Replacing Social Media

I don’t think AI is replacing TikTok or Instagram. That’s still where I discover trends. That’s where creators introduce me to products I’d never search for on my own.

AI is replacing something else:

It’s replacing the conversation that happens after discovery.

The part where you ask:

• “Should I actually buy this?”

• “Which shade would look best on me?”

• “Where’s the cheapest place to get it?”

• “Is there a bundle?”

• “What do you think?”

Instead of bouncing among Google, retailer websites, Reddit, TikTok comments, and group chats, I’m having one continuous conversation with an assistant that already understands my preferences.

The shopping journey that used to look like this:

TikTok → Google → Reddit → Sephora → Amazon → Buy

is starting to look more like this:

TikTok → AI conversation → Purchase

That might sound like a small change. I think it’s one of the biggest shifts happening in commerce.

What This Means for Brands

For years, brands have competed for clicks.

Very soon, they may also compete for recommendations from AI assistants.

If consumers increasingly rely on AI to answer “Should I get this?”, then winning their attention won’t be enough.

Brands will need to provide the information, context, and trust that help AI confidently recommend their products.

The customer will still make the final decision. Brands just need a seat at an increasingly competitive table.

Back-to-School Looks Different Now

Back-to-school shopping has always been one of the biggest shopping seasons for college students.

What’s changing isn’t what we’re buying. It’s who we’re going to for advice.

Gen Z still discovers products from creators, influencers, and friends. But instead of opening ten tabs and piecing together information ourselves, more and more of us are turning to AI to answer the questions that come next:

• Does this fit my vibe?

• Is it worth the price?

• Where should I buy it?

• What goes with it?

For years, those conversations happened with our moms, roommates, best friends, or the employee at Sephora. Now, more and more often, they happen with Chat.

Today, AI is helping us decide.

Tomorrow, it may compare retailers, build the cart, apply the coupons, and complete the purchase on our behalf. If that’s where shopping is headed, agentic commerce won’t feel like a sudden revolution.

For my generation, it’ll simply feel like the natural next step in a conversation we’ve already started.

Subscribe now

Subscribe now

Chuck Gahun is the VP and a Principal Analyst at Forrester where he covers content management technology, content strategy, and commerce services and strategy and co-leads Forrester’s agentic commerce research with fellow analyst Emily Pfeiffer.

He brings about two decades in digital content and commerce, product information management, and tech strategy, and he’s been at Forrester roughly four years. On this episode he’s speaking to the Q2 2026 “State of Agentic Commerce” report.

Listen/Watch Our Interview:

This one's a research deep dive. Forrester just published "The State of Agentic Commerce" for Q2 2026, a client-only report built on a fresh consumer survey, and its headline read like a splash of cold water: the hype is running past reality.

We invited principal analyst Chuck Gahun on Retailgentic to defend that framing, walk through Forrester's (strict) definition of agentic commerce, and separate what consumers are actually doing from what the market keeps promising.

Highlights

• Shopping vs. buying: Forrester defines agentic commerce as “shopping or buying experiences that include AI agents acting autonomously on behalf of consumers, buyers, or sellers”, and until buying is involved, they call it conversational commerce, not agentic commerce.

• The 38 / 17 gap: Among consumers familiar with answer engines, 38% use them for product discovery but only 17% complete purchases, a trust gap Chuck says is the single thing holding agentic commerce back.

• Who’s liable?: When something goes wrong on an answer-engine purchase, 44% of consumers said they’d hold the answer engine itself liable, a sign of how many guardrails, policies, and return paths still have to be built.

• The Amazon corollary: “On Amazon, I’d let my kids, my friends, my family have my credit card and go shopping.” That same trust does not yet transfer to ChatGPT or Gemini, closing that gap is the whole game.

• The merchant verdict: “We’re not telling anybody to wait. We’re telling them to go in strategically while the commerce trust angle is catching up.”

• Machine advantage: Forrester’s four-quadrant merchant framework (agentic fit, machine advantage, operations, and organizational adaptability) pushes brands to build content strategy for machines, not just humans.

• The SKU mountain: The hardest conversation in the field, “you need to upgrade 100,000 SKUs,” and the merchant answers that one SKU takes six months, so “this is gonna take infinite years.” Protocols like OpenAI’s ACP and Google’s UCP keep changing underneath them, on an exponential curve.

• Old B2B PDFs are gold mines: Those spec-heavy, JavaScript-free product pages everyone mocked turn out to be exactly what agents want, and Chuck says B2B (business & industrial) is actually taking off faster, with a report prediction that 20% of B2B moves to human-authorized, agent-executed transactions by year-end.

• Death of the retail website?: A companion Forrester paper argues the website isn’t dead, it’s becoming a repository of data for agents. By 2030, Chuck expects “dual optimization”: machines bring the shopper, and the site’s only job is to convert them.

• Less traffic, 3X conversion: Organic traffic is down, but answer engines filter the traffic so well that conversions are up.

A sharp, honest look at where agentic commerce actually is right now: the trust gap, the merchant playbook, and a few 2030 predictions worth arguing about. Enjoy the conversation

🎧 Listen wherever you get your podcasts: Spotify, Apple, etc.

Or, we have a video version on YouTube (make sure to subscribe!) →

Want to Learn More?

Stay on top of what’s happening in agentic commerce, from the tools powering it to the founders building with it by following us across platforms:

• Podcast: https://retailgentic.transistor.fm

• YouTube: https://www.youtube.com/@Retailgentic

• Instagram: https://www.instagram.com/retailgentic/

• X: https://x.com/retailgentic

• TikTok: https://www.tiktok.com/@retailgentic

We're here to make sense of what’s next in retail + AI and bring you along for the ride. Subscribe and stay tuned!

Welcome to Part IIB of a III Part series: on ‘What’s Next in Agentic Commerce Optimization (ACO)’. Here’s the quick guide for this multi-part series

• Part I: The Next Phase of Agentic Commerce Optimization: Context Capture is here.

• Part IIA: “Product-Level Context Capture: Where, How and When?” is here.

• Part IIB: “Recursive Context Capture Loop Tactical Example” <YOU ARE HERE>

• Part III: Introducing ACO Everywhere 🚀 is here

Recap of Context Capture

In Part I, we introduced the idea that the next big unlock in Agentic Commerce Optimization is product-level context capture. We looked at how the Answer Engines are telling us they want this context by opening up 4 massive high-speed lanes of data (inside UCP/ACP) to provide this unprecedented level of context. Then in Part IIA we looked at where you can go to capture the context. Finally we introduced the concept of a Recursive Context Capture Loop (RCCL).

Part IIB: Tactical Example of SKU-level Context Capture

So far in this series, we’ve been dealing in the theoretical, and now it’s time to get ‘real’ and deal in real-world examples. As I write this, it’s raining and my favorite raincoat is this Columbia Watertight II jacket, so i’m going to use it as an example because I’m familiar with it and long-term readers are probably burned out on dog treats ;-)

Disclaimer: I picked this product at random, as part of this experience, I’m going to purposely try to find contexts of all kind including both positive and negatives. This is part of the context capture experience for every SKU and we always recommend looking at both sides of the bell curve for contextual clues. In fact, it’s one of my favorite products.

Answer Engine Contextual Clues

Since this is Retailgentic, let’s look at the Answer engines first. To save time, we’ll look at the top 2: ChatGPT and Gemini.

ChatGPT

Lower on the offer card, there’s a ‘What people are saying’ section:

Here’s where we find our first contextual clues:

ChatGPT Context pros:

• Better value in entry-level jackets

• Ability to keep rain out

• Packability

• Comfortable fit

ChatGPT Context negatives:

• Less breathable

• Heavier than other options

Pro Tip: Always ask an answer engine specifically for negatives:

I won’t list the 5 negatives here, but I do want to point out the products that ChatGPT does recommend as alternatives, these are always good to know: Patagonia Torrentshell 3L, REI Rainier, Marmot Precip Eco, Columbia Hikebound II.

Gemini

Gemini has very helpful offer cards - as you go down the offer card you get some really good context. For example, the ‘What to know’ section here:

Note that we’re on step 2 and we’re still finding different pros and cons - an example of the

Then even further in the review section we get more→

Google is surfacing right here for us the top questions it receives on this product.

Finally at the bottom of the offer card, it surfaces up reviews of interest. It surfaces these two negative Reddit reviews in that section:

Unfortunately Gemini summarizes these as: “doesn’t hold up long” and “pockets don’t protect phone”.

Summary Context from Answer Engines:

Contextual pros:

• Good value in entry-level jackets (2x)

• Ability to keep rain out (2x)

• Packability

• Comfortable fit

• Customizable

• works well in wind

Contextual negatives:

• Issues with hood fit

• Zipper durability

• Sizing

• Pockets leak on phone (reddit)

• Durability (reddit/gemini and ChatGPT)

• Less breathable (repeated as clammy)

• Heavier than other options (ChatGPT+Gemini surface)

• Not for super heavy downpours (2 layer not 2.5L/3L)

• Warmth - shell, not insulated

Other contextual clues:

• FAQ: Does it have pit zips?

• FAQ: What’s the breathability?

• FAQ: How often does the waterproof coating require reapplication?

Retailer Agent Context

This item is not available on Walmart, so let’s look at Amazon. Here’s the PDP on Amazon. On Amazon we always first look at the Alexa prompt pills. These live in three places:

• Top of PDP

• Middle of PDP Q+A section

• Alexa left-pane invocation

In this screen shot I’ve gathered all three:

Here’s the union of these:

1. Is this jacket machine washable?

2. Does it have a hood?

3. Can it be packed into a small bag?

4. Is it good for hiking?

5. Compare with similar

6. Ask something else

7. Why you might like this

Let’s pause for a second here. Amazon has the largest catalog and is the largest retailer. They are also further up the learning curve on Agentic Commerce than any other retailer. They would never publicly talk about it, but you can bet they are on version 3.0 of their own RCCL and this is is one of the outputs. The Amazon context engine is telling us these 7 questions are the most frequently asked questions/concerns have about this SKU.

One of the most impactful for brands to see is the ‘compare with similar’ - here, again, Alexa is quite mature data-wise and ACO-wise, it is going to show you what Amazon’s customers are most frequently comparing to with this SKU - competitive context is very helpful. Unfortunately it does it by making the user scroll across a panaromic comparison grid. We’ve unfolded it for you here→

At the top Alexa is showing the anchor product, then it gives us two rain coats at a higher price point (MARMOT and North Face) and one at a lower (Eddie Bauer). Notice the comparison criteria - also very important context:

Waterproofing, breathability, Pit Zips, packable, Weight

Also there’s a ‘dog that isn’t barking’ here - why isn’t Amazon offering a higher-end Columbia product instead of competitive brand - it frequently does in other scenarios. Hmmm 🤔

Website Context

Unfortunately, we don’t have access to Columbia’s website. But if we did what we’d look at these areas for contextual clues:

• On-site search engine - Columbia has a typical search engine (more on this on a second)

• Personalization platform - Unknown what they are using

• On-site reviews - There’s no easy summary here, but we do have access to them so we’ll run an agent to do this.

• AI Agent - While they have an agent, it requires login and appears less about finding products and more about post-sales customer service queries.

Website: On-site search Context Strategy

For your on-site search engine, you want to look at the head and the long-tail of the distribution curve of search terms that result in PDP traffic and/or conversion of the SKU. At the head, you want to make sure all the high converting search terms are captured and not only integrated into your pdp, but your entire outbound catalog. On the poorly converting terms or the null search results, you need to have an AI agent analyze those for possible misses in contextual clues. Perhaps a new shopper behavior is developing you’re missing. For example, I remember my Grandfather called raincoats - ‘slickers’. In the modern world, all it takes is about 10 minutes for a celebrity to coin a new phrase or be seen in a product and it goes from ‘the Columbia Watertight II’ to “Hailey Bieber’s raincoat’

Website: Analytics Context Strategy

After the search engine, the other very valuable data on the site is analytics. We recommend looking at your analytics by looking at the PDP traffic and putting into four quadrants as illustrated:

In these high traffic quadrants, we want to first look at the low conversions - this is where we’re getting a hint from consumers there is missing context here, because they didn’t convert. What was missing? This is where an on-site agent can give you more contextual clues a bit further down funnel than analytics and search terms, but since we’re looking deep (website) and deep (across answer engines, website, retailer agents), many times we’ve seen people do this exercise and those contextual clues essentially tell them what’s going on.

For example, Amazon told us top questions asked of Alexa are:

• Is it machine washable? (not on PDP)

• Does it have a hood? (on the PDP down the page in ‘details’, but should be clearly stated above the fold - pictures with hood are behind a semi-transparent part of site.

• Can it be packed into a small bag? (again on the PDP, but far down in details)

• Is it good for hiking? (an occasion type question frequently missed by PDPs). The PDP has this: “Uses: Hiking, Everyday”. That’s not enough context and is a symptom of the over optimization of PDPs for keywords/SEO. In the SEO era, SEO experts would tell you not to be too wordy because that would delete the impact of the keyword hiking so we end up with: “Uses: Hiking, Everyday”. How about - “Whether you’re in the Pacific North West, Maine, Florida or all the places in between, if you are on an a warm weather hike this jacket is perfect protection from wind and rain.” It’s Omni-tech technology is both protective and breathable. The size customization makes it great for hikers that like a tight or lose fit and the hood not only protects your head from getting soaked, but provides valuable sun protection.”

After the high traffic, low conversion, we want to look at high traffic high conversion - what is on the PDP in these examples that’s making this page click? Are we taking everything that’s working here, capturing that context and publishing it externally?

Since we don’t have access to Columbia’s site, we don’t have any new context to add from this section. But hopefully Social Media gives us a last boost.

Website: Reviews

I like to use all the AI tools possible, and for this we need an agent that has access to a consumer-level browser so I had Claude cowork w/ chrome do the negatives and Perplexity Comet Assistant/Computer do the positives.

Website: Reviews: Positives

Here’s a summary of several hundred positive reviews focused on the top 5:

1. Keeps you dry — the top positive, cited constantly. People report staying dry through real tests: daily rain across a full Ireland trip, kayaking in Alaska rain for 2 hours, hiking full rainy days on the Appalachian Trail, a Texas tropical storm, yard work in moderate rain. Common phrasing: “kept me dry the entire time,” “water beads up nicely,” “does exactly what it promises.”

2. Lightweight and packable. A huge recurring theme, especially for travel. “Packs down incredibly small in my backpack,” “light enough to push into a backpack,” “compact, lightweight, and keeps you dry.” Frequently called the go-to travel/hiking rain shell.

3. Good fit / true to size. Most say “true to size,” “fits great,” “perfect fit,” with room to layer underneath. (A couple note it runs a touch small and sized up — the mild counterpoint to the reviews’ sizing complaints.)

4. Comfortable and versatile. Praised as an all-weather, everyday jacket — dog walks, commuting, travel, golf, night walks. “Light, comfortable, stylish and weatherproof.”

5. Style, color, and value. Many like the color options (the yellow/”mustard” gets specific love) and the look (”professional look,” “great fit”). Several call out good value and fast shipping, and a number are repeat buyers or bought extras for family.

Website: Reviews: Negatives

And here’s the same treatment for the negatives:

1. Not actually waterproof — by far the biggest complaint (~18 of 40). Over and over people say they got soaked, often in light or moderate rain within minutes. Representative: “leaked everywhere,” “back seam across the shoulders leaks,” “got soaked in a light rain,” “not even CLOSE to 100% waterproof,” “drenched to the skin both times,” and a paddler describing “catastrophic failure” after 3 hours in rain. Several call it “a windbreaker, not a rain jacket.” A couple also say it isn’t breathable.

2. Cheap, thin, noisy material. The most repeated phrase is that it “feels like wearing a trash bag” / “plastic bag” / “aluminum foil.” Others: “very thin,” “flimsy,” “like a thin trash bag,” “cheap looking material.”

3. Fit / sizing problems. Runs small and cut too short; inconsistent sizing between Big and Regular; sleeves too long; “about a size and a half too small”; doesn’t fit over a layer.

4. Build-quality defects. Zipper caught and broke on first use, a hole next to the zipper out of the box, hood elastic snapped on first use, unevenly-sewn/shallow pockets, smeared heat-transfer logo (vs. embroidery), velcro instead of snaps.

Website: FAQs/Q+A

Columbia has a FAQ section which is also available to a browser agent, so here’s what that surfaced:

• Weight: ~0.95 lb.

• Waterproofing/DWR: described as Omni-Tech fully seam-sealed, “so you stay completely dry”; for re-treating over time they recommend Nikwax (TX works well) or GEAR AID.

• Pockets: two zippered hand pockets.

• Layering: regular fit with room to layer; works as an outer shell over most layers.

• Warmth/temperature: no insulation or warming tech — layer for cold.

• PFAs: fabric does not contain PFA “forever chemicals.”

• Packability: packs into its own hand pocket.

• Climate control sheet: directed to contact Customer Care.

Social Media Context

Social media is a goldmine for context. We always recommend looking for your product’s mentions or PDPs on TikTok, Reddit, Pinterest and Instagram and capture all the hash tags, categories, groups, and any influencers that are mentioning your products as the most valuable context.

For our Columbia jacket, some interesting findings:

• There’s a movement called Gorpcore that started out a bit fancier and higher end where urban dwellers wear hiking gear in the city. Today it’s heading to ‘normcore’ or “quiet outdoor” or “technical fashion for the city”. It turns out our Columbia Watertight jacket is popular with this community.

• There’s a bunch of communities under the umbrella “Cool Girl/Cool Girl Hiking” term that encourage females to get out and hike. These communities love gear and our Columbia Watertight (the female edition) is a popular product in this community.

If we had access to the search terms, these types of social-media findings are unfortunately almost never captured, and then fed back to the PDP and therefore will show up as the top null search results.

Social Media Context Captured

To summarize for social media, this was a fast tour due to time constraints, if we had more time we’d spend a lot more time on YouTube and TikTok. But we did find some valuable information already:

• Gorpcore trend

• Normcore

• Cool Girl Hiking

Other Sources of Context: Store and Manufacturer

Columbia has about 200 Columbia stores and says they have 1,850 retail partners. That’s a very rich surface area with humans selling Columbia products every day, answering questions, understanding occasions and the lifestyles where Columbia’s customers use the products. That data and information is gold, and figuring out a process for collecting, collating and feeding that into the Context Capture engine is well worth the effort.

Even though Columbia itself is the manufacturer, we’ve found that even branded manufacturers have a significant loss of valuable information as products go from idea→design→prototype→iterate→final product→scale. Frequently the digital team is given 4 bullets to work with and hires copy writers to recreate copy that was lost in the process. Agentic Commerce Optimization can hopefully become the value creation engine that tears down these silos and enables the digital team to have

Putting it all Together

For our Recursive Context Capture Loop, in this post, we got our hands dirty and illustrated steps 2 and 3 at a very granular level for one SKU: (to the extent possible with a randomly selected product using publicly available information)

Through that process, here is a summary of valuable context we captured about this SKU put into 6 buckets:

1. Pros/positives

2. Cons/negatives

3. Top competitors mentioned/compared

4. Social/lifestyle signals

5. FAQS (from around the web and the PDP)

6. Reddit Authoritative Content

It’s important to note that Columbia itself should be able to pull together 3-5X what we can find publicly from online and offline sources (website, store, manufacturing process, etc.). Today you may say that juice is not worth the squeeze, but I think in < 9 months it will be, so if you want an edge, get started on this today.

What’s Next?

In Part III, coming Tuesday July 14th, we’ll conclude the series and put a proverbial bow on it by tying it together at the strategic level and then come back to our Columbia Watertight jacket and implement what we’ve learned here.

Sparky Eating the Walmart Search Engine..

Back on June 9th, we were the first to report that Sparky had moved from a second-class citizen on mobile-only a full web integration. Today we discovered that in mobile, Walmart has this new experience:

Here you see the July 7th search experience on the mobile app for “Men’s running shoes” before I could get to running, it popped up not only the traditional auto-complete suggestions but a new “Ask Sparky” section. When you click on each of these, it runs a prompt in sparky as if it came in from any other surface.

Look at the % of the page here taken up by Sparky 🤯

What Does This Mean?

Amazon’s adoption of Rufus/Alexa went through a process where it wasn’t good, it was minimized, then suddenly something happened and Amazon went from hiding it to placing it everywhere and they’ve fully integrated Alexa in the search bar. From public comments and reading the body language, we can infer that, just like Answer Engines, the Agentic conversions went from ‘lower than non-agentic’ to ‘equal to non-agentic’ then to ‘oh wow!!!, this is much better than non-agentic’. It feels like Sparky has entered that era.

As a top online retailer, you don’t throw something into the search stream of hundreds of millions of users unless you have an economic reason.

For merchants, it means that, just like at Amazon, where Agentic Commerce Optimization was something you could slow roll, if it’s July and we’re seeing this, by Oct/November, what % of search terms will be routed to retailer assistants? If the conversions are strong and continue, I would guess a LARGE %. This has moved from hobby to core. Time to prioritize ACO on Walmart and Amazon if you are a 3P seller or a branded manufacturer selling wholesale, the time has come.

This does NOT mean that you should give up good retailer-level SEO - that’s still the majority of traffic and our VP of Research has drilled into my head that strong-SEO is the foundation of ACO on both Alexa and Sparky.

What do the prompts say about your category? Tip: This is an important contextual clue that should be captured and fed into your strategy!

What’s Next?

I can’t help but wonder, wouldn’t that be a GREAT place to put a sponsored prompt?

Stay tuned.

Here’s some more examples as a bonus:

Subscribe now

🇺🇸 Editorial note: Due to the Holiday week, we’re going to flip our schedule and get you this week’s podcast early (for your holiday travel ✈️ 🚋 🚗 enjoyment) and if time allows do an Agentic News post this week (for your Holiday Reading 📖 Enjoyment). Enjoy the pod and safe travels everyone! 🇺🇸

Jennifer Peters is the Director of DTC, MarTech, and Digital Compliance at OLLY, the supplement brand best known for its gummy vitamins and distinctive square bottles. OLLY is part of Unilever, and Jennifer joined a little over five years ago, coming on mid-COVID, after the acquisition.

She's a self-taught Shopify developer who got tired of hearing "no" and "that'll take six months," and on the side she still re-platforms brands off legacy systems onto Shopify, which she calls "the democratization of e-commerce."

Listen/Watch Our Interview:

This was a true cold-outreach episode as we never met before. I was reading an article about agentic commerce best practices, and Jennifer's name kept coming up as the executive actually doing the work and sharing it openly. So I took a shot, invited her on, and she said yes.

We start with a Shopify “side quest”, how a self-taught developer ended up in AI god mode with Claude and MCP, before getting into what OLLY actually is and why a brand that does most of its revenue in-store suddenly cares so much about its website.

Jennifer has been quietly building one of the most grounded, real-world Agentic Commerce playbooks we've heard on the show. She shares the basics done tightly and gives a clear-eyed take on what's worth investing in while the ground shifts week to week.

It’s a grounded, refreshingly un-hyped conversation about building for the agentic era without losing your head, or your customers. Hit play.

Highlights

• The cold-resource, high-impact win: Jennifer on catalog feed enhancement.

• The FAQ slam dunk: OLLY added FAQs to its PDPs using content already on the page, just repackaged into bite-sized chunks. LLMs love the format, and so do humans.

• AEO is just good SEO: Clean up your schemas, clean up your data structure, and do the things that are good for your brand anyway.

• Attributes Google never asked for: A partner ingests OLLY’s public catalog and pushes back AI-ready attributes, none of them customer-facing.

• The website is now the point of entry: A site that was “7 or 8% of the business” is now the brand authority source agents go to first, for OLLY’s own site, Walmart’s Sparky, and Amazon’s Rufus.

• Trust but verify: OLLY triangulates recommendations across Profound, Contentsquare, and Envive. If all three flag the same thing, it’s probably worth acting on.

• Don’t “do AI to do AI”: “We’re not changing the goals because we have AI. We’re keeping the goals and just doing a better job getting to them.” Same OKRs, just faster.

• Fix the un-AI basics first: Connect your DAM, your creative stack, and your marketing stack before you try to sprinkle AI magic on top. Scot adds that AI can be the excuse to finally get it prioritized.

• The AI bubble in the bubble: From San Francisco, Jennifer on the coming counterculture: “No AI in our creative” as a selling point, growing data center backlash, and Ronny Chieng’s Harvard commencement speech.

Jennifer is exactly the kind of hands-on practitioner this show exists to find: strategic, tactical, and refreshingly honest about what’s real and what’s noise. We’ll have to get her back in a year to see what other magical (and weird) things she’s figured out.

Enjoy the conversation.

🎧 Listen wherever you get your podcasts: Spotify, Apple, etc.

Or, we have a video version on YouTube (make sure to subscribe!) →

Want to Learn More?

Stay on top of what’s happening in agentic commerce, from the tools powering it to the founders building with it by following us across platforms:

• Podcast: https://retailgentic.transistor.fm

• YouTube: https://www.youtube.com/@Retailgentic

• Instagram: https://www.instagram.com/retailgentic/

• X: https://x.com/retailgentic

• TikTok: https://www.tiktok.com/@retailgentic

We're here to make sense of what’s next in retail + AI and bring you along for the ride. Subscribe and stay tuned!

• Part I: The Next Phase of Agentic Commerce Optimization: Context Capture is here.

• Part IIA: “Product-Level Context Capture: Where, How and When?” <YOU ARE HERE!>

• Part IIB: “Recursive Context Capture Loop Tactical Example” is here.

• Part III: Introducing ACO Everywhere 🚀 is here

In Part I, we introduced the concept of context capture and how it is now necessary to beef up the product-level context to match the sharp increase in shopper context.

Product Catalogs in the Agentic Commerce Era

Before 2022, Product Catalogs were largely static. There were new products added every quarter and some products ‘deleted’. Seasonal businesses like fashion had more movement like that. But once created, each individual SKU didn’t really change except for price and inventory which have been increasingly dynamic over the decades.

In the post Agentic Commerce Era, due to the new importance of integrating context into product catalog as covered in Part I.

Where Should You Capture Context?

There are 6 major areas where you can capture extremely valuable consumer shopping product-level context. They are:

1. Answer engines - With Google releasing AI Performance insights and rumors that ChatGPT’s ad engine is going to be sharing a similar level of information. We are entering an era where Search Engines shared little to nothing, but Answer engines are going to have much more transparency. This information is valuable context capture.

2. Retail agents - Like Answer Engines, I suspect that Amazon/Rufus, Walmart/Sparky and other retailer shopper agents like Target’s TSA are going to be providing more infromation about what prompts shoppers are using, the topics being discussed, product features compared, comparative products, etc.

3. Physical stores - If you are omni-channel, your retail stores are a wealth of context that you previously haven’t thought of leveraging elsewhere, but now in the AI era, there’s very rich information you should vacuum up for your on-line efforts. Sales rep training manuals, product guides and anything that a store associate would be trained on can be leveraged by an LLM for product-level context.

4. Brands/Manufacturers - As a retailer, the deepest level of product information is typically available from the manufacturer, and then frequently this is either ignored and recreated or condensed down to 2% of the original content. If you are the brand, frequently the online group actually does the same thing as a retailer - takes the design briefs, hundreds of pages of product information, spec sheets, technical details and boils it down to 1% for the human readable PDP, which is normal, but now we live in a world where all of that information is gold dust and we need to go figure out how to keep it at 100% from offline→online.

5. Social media channels - For hundreds of years, centralized merchants were the taste makers in retails, but we are now in the Kardashian/TikTok era and the tastemakers are independent influencers, celebrities, athletes and content creators. When influencers recommend, are seen with or even have a similar product to something you offer (either paid, organic or collab’ed) , that is important context that needs to be captured.

6. Website - Your online website is one of the largest surface areas of lost context. Each shopper is browsing, searching, receiving personalizations, leaving and consuming reviews and increasingly prompting your on-site shopping agent. Those contextual artifacts on both the positive (high converting) and negative side (low converting) are significantly valuable contextual clues to product-level shopper behaviour that needs to be captured, gathered and actioned.

Across these six opportunities to increases context capture, every retailer will have different amounts of context and priorities for their business. In fact it can seem overwhelming when you really start thinking about how much context could be captured. We recommend this best practice to solve that problem:

Building a Compounding Context Capture Recursive Loop

Before AI, workflows across programming/coding and optimizing digital sales be it ecommerce or ads or what-not was a mostly linear process with a bit of a maybe quarterly, semi-annual or annual closing of the loop. Since November 25 when the agentic capabilities of the ChatGPT and Claude models took a huge leap forward, developers have realized that with AI automating a lot of the process that these ‘loops’ can be significantly tighter. Developers are particularly obsessed with these loops now.

Why are AI Developers Obsessed With Loops?

There are two types of Loops - ones with AI in the middle as the ‘judge’ or ‘tagger’ and then those without a human.

If a human is in the middle it’s called Reinforcement Learning with Human Feedback - RLHF. If AI is in the middle, it’s called Reinforcement Learning with AI Feedback. As you can imagine, relative to AI, humans are slower and less scalable and more expensive so if RLAIF is » RLHF. (Want to learn more? Start here.)

This started with some basics and then one of the most famous AI developers, Andrej Kartpathy who has a mini llm he calls nanoGPT, created a RALAIF recursive loop so this llm model codes itself and if the new version improves it’s benchmarks, it is good, otherwise, it deletes that and tries again. In this post, he ran it in this mode for 2 days on this loop→

In this chart down and to the right is better. You can see that after 2 days with zero human interaction the sytem got better.

Today, ChatGPT and Claude are built this way - they build a new version of themselves. If that version is better. v2 then builds a better version to v3 and so on. This is why model releases are bending the exponential curve. (Anthropic hired Andrej btw).

While that’s a little scary 🤯, let’s stay in our World where YOUR job is to drive more sales and profit for your org (while crushing the competition) and answer the most important question of the day:

What Do Recursive Programming Loops Have to Do With YOU?

This concept of loops can be applied to any system that has a definitive zero sum or measurable outcome. That’s not only coding, it’s digital retail and marketing.

• Example from coding: Did the code compile? Did the code pass the tests? Was the desired outcome generated by the code?

• Example from Digital Retail: Did the SKU map correctly? Make a change to optimize the product catalog - did that that increase conversions? Did the ad hit the ROAS target for SKUX?

• These are perfect for the digital vs. analog world because in the digital world no humans are in the loop so the loop is faster and tighter. For example, if you wanted to use RLAIF for store inventory, the loops are too slow, over 3yrs it will have an impact, but not 6 months.

I believe that in the next 18 months we’re going to see significant digital retail disruption due to these loops. Some forward-leaning retailers and brands, are going to lean into these and use them for digital advertising.

Answer engines (Claude, ChatGPT, META, SpaceXAI, Gemini) are going to be using these to hyper-improve the customer-experience and conversion rates of Agentic Shopping and the performance of AI Ads.

Subscribe now

Quick Recap to Tie it All Together (so far)

If you’re with me, here’s the logic path from Part I to where we are here at the end of Part II:

• Product-level Context is the next frontier - After doing the basics (steps 1-5 of our agentic commerce optimization framework), and the advanced step 6. To do steps 7,8,9 you need a loop (that’s why we put those little circles there back in Sept of 25 ;-) )

• UCP/ACP have given us plenty of ‘room’ for context - In addition to much wider basic attributes and conversational attributes, we have an essentially unlimited product-level FAQ data feed we can send the answer engines→

Good news! We now have all the pieces in place to build such a system….

Leveraging the AI Loop Concept and Context Capture: Compounding Context Capture via a Recursive Loop

I believe what all retail and DTC brand CDOs need to be thinking about maybe by Holiday 26, but definitely by the end of 2027 is how to create a recursive compounding context loop or a RCCL. Here’s what it looks like:

Step 1: Optimize Product Catalog SKUs

In this step we publish the latest version of the product catalog and measure the impact. While that measurement is happening we simultaneously collect context.

Step 2: Gather Online Context (Answer Engines, Website, Socials)

I grouped the online sources together because being digital they are always going to be faster than offline. You can API/MCP into your datasources or larger orgs will have a real-time-ish data lake they can pull from. Because of this maybe you pick up digital every time through the loop and offline ever 4th time or maybe once a month type thing for what fits your org.

Step 3: Gather Offline Context (Store, Manufacturer)

Offline is always slower and won’t update as much as online, so maybe for your org once is enough through here, but if you have stores, you should prioritize a more real-time way to pull store associate feedback- specifically any interesting consumer behaviors changes/nuances into the system. That can be made pretty quick, especially if you are on a retailing type in-store system.

Step 4: Evaluate Context and Prioritize

In Step 4, the evaluation from the current version is done and opportunities for improvements have been discovered and made. At the same time, new context from steps 2 and 3 is pulled in and evaluated. A new version of the catalog is ready for the next loop.

Step 5: Loop with new context in the Product Catalog

Catalog new version is ready to go and published and the loop starts again. Faster and faster.

I realize this is significantly different than the way we have operated in the past and it maybe seem like overkill, and we’ll get to more of why this is a priority in the next post.

What’s Next?

I was going to keep going right into examples here, but that’s a lot to digest and sit with, therefore I’m going to split part II into 2 parts IIA and IIB. Part B will be out July 8th and Part III will be out on July 15th. Until then we’ll continue to have awesome pods for you and also don’t forget to subscribe to our new channel! - www.retailplaybook.ai

Have a great Independence Day everyone! Here’s to 250 years of America! 🇺🇸

Subscribe now

Here at ReFiBuy/Retailgentic HQ, we talk to hundreds of brands and retailers every week. Amongst brands, both Hybrid (DTC+Wholesale) and Wholesale-only Brands, as well as large 3PM sellers, one theme has gone from a whisper to a roar:

We’ve talked to top brands who have been top 1/2/3 in their category for years that are now seeing disruptors leverage Rufus/Alexa to upsurp them. We’re seeing dynamic and nimble 3P sellers who understand the importance of this on Amazon and Walmart surge in sales while incumbents scratch their heads as they look everywhere to understand what’s going on.

These Retailer agents are very new and all built on LLMS so we’ve seen some pretty crazy hallucinations like saying something is Kosher when it isn’t or telling a shopper that some chicken nuggets are vegan (they are not).

The Hunger for more “Retailer” Answer Engine Content

As these conversations have surged, so has the popularity of our content on Retailgentic on this topic. Our coverage of the Rufus data from SensorTower here and here was extremely popular and viral. Our Sparky coverage has been equally well received and the messages from readers like you has been - we want more!

Enter: Andrew Bell

After looking at the data at how popular our Rufus→Alexa and Sparky articles are, we had a wacky idea, what if we had someone that focused 100% on this. We researched everything we could about optimizing systems like Alexa and Sparky, every piece of content pointed back to one person who was the authoritative source of information, largely in the form of playbooks.

Close observers took note when we announced that Andrew Bell had joined as VP of Research and we got a LOT of questions about what was going on.

Well fellow Agentic Commerce Maximalists, the time has come to reveal a bit more about what’s going on, but not everything 😀

Introducing RetailPlaybook

Today, we are launching RetailPlaybook, a new ‘channel’. It’s a substack newsletter and companion podcast authored and hosted by Andrew.

RetailPlaybook’s Focus

Built for 1P Brands and 3P sellers, RetailPlaybook provides cutting-edge research, original heavily researched playbooks, and future-proof strategies to optimize for agentic commerce on Amazon’s Alexa for Shopping, Walmart’s Sparky, Target Shopping Assistant and beyond. As the name implies, Andrew is going to be publishing regular Playbooks that address the current and top concerns from 1P and 3P brands and sellers. The initial focus will be Amazon, but there will be plenty to cover on Sparky and Target as well.

Here at Retailgentic, we’ll continue our focus on the Answer Engines and the occasional big news items over at Amazon and Walmart and the overall Agentic Commerce industry.

Check out RetailPlaybook Today!

Today the inaugural substack article is available here and podcast is live.

🎧 Listen wherever you get your podcasts: Spotify, Apple, etc.

Or, we have a video version of the RetailPlaybook podcast on YouTube (make sure to subscribe!) →

Next Steps…

If you want to optimize your product catalog and drive increased sales on Amazon, Walmart or Target, make sure to:

1. Subscribe to the RetailPlaybook.ai substack!

2. Subscribe to the RetailPlaybook audio pod here.

3. Subscribe to our RetailPlaybook YouTube Channel here.

We hope you enjoy this new offering from Andrew Bell!

I’m going to go through this fast, buckle up, this is our first every Retailgentic DOUBLE FLASH - two, yes two breaking Agentic Commerce news updates in one:

1. Target is live with the Google Gemini UCP-powered cart

2. (while exploring 1, we discovered this am..) Google has turned on Universal Cart in the wild.

Note: All the new Google/Gemini goodies start in AI Mode - so assume I’m talking about that always. (AI Mode: go to google.com - click on AI mode - right side of search bar, make sure your Google now says “AI Mode” in the upper left.)

Quick history here:

• Jan 11 - Google announces UCP at NRF

• Feb 11 - UCP went live

• March 19th - Google announces multi-item cart and Loyalty

• April 22nd - Microsoft DUMPS ACP, Launches USP with Target and first Loyalty integration.

• May 19th - Google announces Universal Cart (their implementation of UCP multi-cart) is done and coming in “Summer”.

• June 19th - The first Retailgentic Double Flash🚨, we have both Target now launching on Gemini PLUS the first Universal Cart live sighting with live demo!

Live Demo

The best way to see this is via this ~10m video. I was doing the Target Cart demo and had tried it on another account to get a feel for it. That account did not have Universal Cart, so mid-demo here it turned into a double demo. A walkthrough follows.

The Trigger: blog post update from Target

What got me here is several Target watchers saw that on socials they mentioned that they had updated this blog post and when you look at the update, it’s saying they are live on Google checkout.

I tried last night and couldn’t get it to pop so checked first thing this am (rollouts take overnight sometime to propogate to all servers/accnts.

Here’s what I discovered this am→

Detailed Walk Through of Target and Etsy in Google’s Universal Cart

Step 1: Research→Find

I skipped research and went right to a sku I knew would be at Target and trigger a product card, but it also showed the new add to cart! (We’ve seen visuals of this from Google I/O and GML).

Step 2: Add to Cart 🎉

When I added the falcon to the cart I got this→

And in the upper right of my google account is now a persistent cart icon! Feels very, umm, Amazony! 🤔

Here’s the little cart that could forever change Agentic Commerce and retail→

Step 3: Load that CART UP!

Since I was full on ‘hallucinating’ (because a universal cart isn’t possible according to the naysayers), I figured heck may as well make it a good one and I loaded the cart with an Etsy item and a Target bag:

Step 4: The Holy Grail 🏆: True Universal Cart Nirvana!

You can see in the above screenshot I have two options: “Buy Now” or “Checkout on Target”

If I hit Buy Now → 2, yes 2 TARGET items in one checkout, the holy grail has been achieved!

OR, If I hit Checkout on Target, it does cart injection and takes me there→

What’s Next?

We’re going to spend the weekend decomposing this and learning more, stay tuned. As I was writing this, Walmart seems to have flipped to ‘on’ as well→

Live on the exponential growth curve, buckle up everyone this holiday is going to be BONKERS!

Subscribe now

Today, Shopify announced their Spring Edition and it’s chock full of Agentic Commerce features. If you want the total details of the release, they are provided here.

Subscribe now

I was able to get an early look at the features for Retailgentic readers and then ask a Shopify representative for some clarifications on the features. We’ve identified four of the biggest features that focus on Agentic Commerce. For each of the four, we first cover it’s high-level capabilities (as briefly as possible, more detail is in the Spring Edition detailed post linked above) and then cover the Q+A provided by a Shopify spokesperson.

The Four Features (out of many more) that caught our eye are:

• Agentic Admin

• Shop Campaigns

• Sidekick App Extensions

• Shop App Personalized Shopping Assistant

Up first, Agentic Admin:

Agentic Admin

Spring Edition Feature: Managing Agentic Storefronts and Insights in Shopify Admin

For the first time, merchants have one place to manage Agentic Storefront channels in their admin, track performance across channels, and see actionable insights. What it unlocks:

• Merchants can now see orders, sales and conversions from ChatGPT, Google AI Mode and Gemini, and Shop at a glance. Products are automatically accessible to AI channels through Shopify Catalog

• Search intelligence shows the top AI queries in a merchant’s category, including which ones they’re ranking for, and where they’re not showing up paired with actionable insights

• Knowledge Base shows merchants which questions AI agents are asking about their brand that haven’t been answered yet (like retail locations, bulk orders, and customer service) and lets them fill those answers in directly from the Admin.

Agentic Admin Q+A

Q: Search intelligence - Google announced AI Performance Insights at GML. This seems similar but across all answer engines. What can a merchant learn here?

There’s a natural comparison but the scope is different. Google’s AI Performance Insights shows how a brand performs across Google’s AI surfaces specifically, including share of voice against competitors. Shopify’s Agentic section in the admin looks across all the AI channels Shopify connects to: ChatGPT, Copilot, Google, and Shop.

Merchants can see the top AI queries in their category, whether their products rank for those queries, and where they’re not showing up. They can search any query against the Catalog API and see the exact same results AI agents see, including top-ranking products across all merchants, not just their own. They can also preview their actual ranking live in ChatGPT and Shop with one click.

Where Google’s tool focuses on visibility and share of voice within Google’s ecosystem, Shopify’s gives merchants cross-channel intelligence plus actionable next steps. If a product isn’t ranking, listing quality signals tell the merchant specifically what to improve: descriptions, images, reviews, pricing, options, policies, or titles. Sidekick can then recommend fixes. So it’s not just reporting, it’s a path from insight to action across every AI channel where their products could appear.

Q: Can you say more about where the data comes from - is this panel data or shared via Google/ChatGPT/etc.?

The search query data comes from Shopify Catalog. When a merchant searches a query in the Agentic section, they’re hitting the actual Catalog API, so results reflect what AI agents see when they search on behalf of buyers. Performance data (sessions, orders, conversion, sales) comes from Shopify’s own analytics, attributed by channel. AI channels don’t currently share impression data with Shopify, so the metrics are sessions, orders, conversion, and sales, not impressions or views.

Q: Knowledge Base recommends questions that aren’t answered - where do these answers go? (e.g. meta-data on the site, data feeds to the engines, on the human readable PDP?)

The Knowledge Base surfaces questions that AI agents are asking about the merchant’s brand that haven’t been answered yet, things like return policies, shipping details, retail locations, or bulk order info. Merchants answer them directly in the Admin through the Knowledge Base feature.

Those answers become the authoritative source that agents reference when buyers ask about the brand. Instead of an agent guessing or generating an answer from scraped web content, it pulls the merchant’s answer from the Knowledge Base. The answers live in Shopify and are available to agents through the same infrastructure that powers Catalog, so agents get facts, not guesses. Think of Knowledge Base is for shops as Catalog is for products. It’s a better alternative to scraping for this information.

Subscribe now

Shop Campaigns - Growing your business with AI-driven campaigns

Shop Campaigns, Shopify’s performance marketing product, expands to three new surfaces: Pinterest, ChatGPT, and the open web via Microsoft Monetize. Merchants only pay when customers convert.

Shop Campaigns Q+A

Q: ChatGPT has rolled out a wide variety of ad formats, bid options, etc. What’s the matrix supported?

We have been experimenting with all formats of product catalog based advertising that OpenAI offers to determine what works best for our merchants. We don’t share specifics of the exact settings or formats that Shop Campaigns operates on, at a given point in time, on any platform.

Q: Is support there for Gemini’s various ad formats?

We advertise on Google across a variety of ad formats and campaign optimization capabilities. We don’t specifically target Gemini ad formats, but some of the ad formats that we leverage may lead to ads in Gemini.

Sidekick App Extensions

Sidekick App Extensions connect third-party apps to Sidekick, so merchants can get AI-powered answers and take action across their full tool stack from a single conversation.

Sidekick App Extensions Q+A

Q: Any conceptual examples of the types of things Shopify would like to see innovated as an extension to sidekick? Any launch partners there?

App Extensions launched with 15+ partners including Klaviyo, Loop, Smile, Judge.me, Matrixify, Avia, Seguno and Checkout Links. Any developer in Shopify’s ecosystem can build one going forward.

Examples include:

• Merchants using Klaviyo can ask “find my best-performing subject lines” and get real campaigns back, with open rates and click rates pulled from live data.

• A merchant using Judge.me can ask Sidekick to “help me respond to recent low-star reviews,” and Sidekick opens the right page in the app with the review context and a draft reply already filled in. The merchant reviews and approves before anything posts.

Q: Is this the same framework and business model as the app store (15%) or is it more like an open MCP protocol type thing?

There is no additional cost for developers or merchants who use Sidekick app extensions. Developers configure their extension and merchants benefit from Sidekick’s ability to access data and actions from their apps.

Shop App Personalized Shopping Assistant

Shop App, Shopify's consumer shopping app, now has a shopper-facing personalized shopping assistant built into search. Buyers describe what they are looking for in plain language and get recommendations shaped by their real purchase history, browsing, and stated preferences. Each interaction sharpens the recommendations over time.

Shop App Personalized Shopping Assistant Q+A

Q: You’re announcing a personalized shopping assistant in the SHOP app - will this PSA be available for merchants to put on their store so they can have their very own ‘rufus or sparky?’

Agentic search on Shop is specific to the app and web that buyers can use throughout their shopping journey. Buyers can tap the search icon on Shop and ask it any question in natural language. Buyers can use search to discover and purchase products, ask about a merchant’s return policy and shipping updates, and more – all directly on Shop. Search learns a buyer’s taste over time and asks smart follow-up questions to provide the most personalized recommendations.

Shop search is built on Shopify’s commerce infrastructure — the Shopify Catalog, Shop Pay identity, buyer profiles, and MCP tools for cart, checkout, order lookup, and more.

Subscribe now

Roger Dunn has watched retail media grow up from the agency side (Mediacom, GroupM, WPP), the tech side (he launched Criteo's retail media business in Australia), and the brand side (a global retail media role at Diageo, home of Johnnie Walker, Guinness, and Baileys). Now he's CCO at Thrad, a Silicon Valley startup building the ad infrastructure for LLMs, "ChatGPT ads, but for everyone else."

Listen/Watch Our Interview:

In this episode, Roger and Scot get into where advertising actually lives inside answer engines, what's working today, and how brands should think about showing up when shopping stops looking like a search box.

Highlights

• What Thrad is building: both the supply side (helping retailers and chatbots monetize conversational surfaces) and a self-serve buy side where brands can run ads across existing AI inventory globally, not just in the four countries where ChatGPT ads have launched.

• The new ad formats: sponsored messages, sponsored prompts, polling ads, and product carousels, plus the steady stream of formats Google unveiled at Google Marketing Live.

• AI-native campaigns: how Thrad turns a URL, a brief, or a prompt into keyword, prompt, and dynamically generated persona targeting, and why creative (human faces beat a bare logo) is where the human edge still lives.

• What ChatGPT ads look like right now: CPCs roughly in the $1–$5 range, and why the early “hot mess vs. holy grail” takes are both missing the point.

• Roger’s definition of agentic commerce: not a robot doing the final click, but AI shaping every stage of the journey: discovery, comparison, price, review synthesis.

• The state of the field: Gemini’s full-court press, Perplexity’s Comet browser, Meta routing UCP checkout into Instagram, the Shopify question Wall Street keeps asking, and where Anthropic/Claude might (or might not) play.

• Down Under as a test lab: Woolworths’ Olive, Bunnings’ Buddy, a duopoly retail market, and why Australia keeps getting the pilots.

• The advice brands actually need: start small (think $10–20K in a market) but start; get out of “keyword jail”; and the analogy that lands it: keywords are ice cubes, prompts are snowflakes. A few uniform ice cubes everyone fights over, versus infinite unique prompts where a challenger brand can finally own its wedge.

This is a genuinely useful tour of the side of agentic commerce most of us squint at, ads, from someone who has built the playbook three times over. Enjoy the conversation.

🎧 Listen wherever you get your podcasts: Spotify, Apple, etc.

Or, we have a video version on YouTube (make sure to subscribe!) →

Want to Learn More?

Stay on top of what’s happening in agentic commerce, from the tools powering it to the founders building with it by following us across platforms:

• Podcast: https://retailgentic.transistor.fm

• YouTube: https://www.youtube.com/@Retailgentic

• Instagram: https://www.instagram.com/retailgentic/

• X: https://x.com/retailgentic

• TikTok: https://www.tiktok.com/@retailgentic

We're here to make sense of what’s next in retail + AI and bring you along for the ride. Subscribe and stay tuned!

👀 First Look at Claude’s New Agentic Commerce Product Cards 🎉

Our ReFiBuy monitoring agents are constantly running product-related prompts across all the Answer Engines. Yesterday, very unexpectedly they started reporting that Claude is now showing product cards. Here’s an example for a hairdryer:

It’s important to note that these Claude Product Cards are pervasive across all categories. Here’s a collage of a variety of product cards across several categories from electronics to pet to fashion and beauty:

While these are rudimentary, the intent is clear here. If you recall back on Feb 9th, we pointed out that Claude had mentioned they are “particularly interested in the potential of agentic commerce, where Claude acts on a user’s behalf to handle a purchase.”

In that same piece we pointed to reports from NRF in January that Anthropic was meeting with Merchants and we’re aware of representatives that were at Shoptalk also holding highly confidential meetings.

Historically, over the 27yrs I’ve been working in retail, this time of year is when new features are frequently introduced with a ramp into the Holiday selling season. Could we see Claude rapidly go from ‘Research→Find→Buy’ in time for Holiday 26?

Updating The Five Levels of Agentic Commerce Tracker: Claude ‘up’ and Perplexity ‘down’.

Here’s our updated Agentic Commerce Levels of Autonomy tracker:

As the arrows indicate, we are downgrading Perplexity to Level-1: Research. Not only has Perplexity removed the buy capability, but according to our Monitoring agents, they have eliminated all Product Cards. Optically Perplexity seems to be focusing on the financial vertical and their Computer agent (kind of an OpenClaw type user agent with more rails that OpenClaw).

At the same time Perplexity is moving down, we are now upgrading Claude from Research to Find due to their first iteration into Product Cards.

That yields this end-state for where we are today on the tracker (minus the arrows in case you want a clean version).

Note you are free to use this in your presentations, we appreciate attribution for Retailgentic and a link back if you do publish it online.

Subscribe now

That’s not the only big Agentic Commerce news today→

Adobe Releases Q3 AI Traffic Report: AI-Driven Retail Traffic is up a Startling 138% Y/Y!

The first datapoint Adobe dropped is that Retail sales accelerated to 138% y/y growth (the green line above). That’s impressive, but check out the yellow line! AI driven travel spiked 194% y/y. Financial services spiked to 105% y/y as well. My educated guess is as Gemini and ChatGPT - the big 2 from a traffic-perspective are ramping up their ad programs, they are working on the user experience, perhaps adding touches like product cards and other elements that are accelerating this move.

Also note that going back to October 2024, AI Driven Retail Sales are up 1,324%! obviously it’s a small base, but at 138% it’s not slowing down, it’s accelerating!

On top of that, this is not ‘garbage firehose traffic’→

Look at this trend, as Answer Engines are building out their training set, building up memory capabilities and understanding Shopper Intent, and getting better at Product-level information - the ‘match rate’ is growing rapidly. Today swirl all that together and visits initiated by Answer Engines are worth 53% more.

The Adobe Q3 Report offers two insights into why:

On top of consumer sentiment improvement (this is from a n=3,000 survey performed in March 2026), Adobe has some pretty crazy data that is the first statistic I’ve see on the topic of Agentic Commerce and returns:

There’s a lot of scary aspects of Agentic Commerce:

• Rapidly changing traffic patterns

• Increasing evidence that Google is accelerating the path to zero clicks

• Lot’s of hard work to upgrade your catalog that can seem insurmountable

• An alphabet’s soup of protocols

• New players entering and leaving the arena at a frenetic pace (see above)

But what if Answer Engines are a bit of a Holy Grail? Evidence is mounting to support this:

• The traffic converts 54% higher (the real-world data I can see points to significantly higher than this believe-it-or-not)

• The AI Ads are significantly more performant than pre-AI Ads

• And now…Agentic Commerce Traffic could reduce returns 69% 🤯

Finally, while Google is acceleration the decrease in SEO/SEM traffic, Answer Engines growth is accelerating. I believe the ‘traditional search’ and 'Answer Engine Search’ lines are going to cross faster than anyone ever predicted.

Adobe’s Updated Demographic Information: Zillennials Are Driving the Agentic Commerce Trends

These types of adoption curves are common - in the early days of E-commerce it was the same thing. What’s new, is this curve is more pronounced. That delta between 53% at Gen Z and 34% at Baby Boomers is a 19% difference - huge.

Where to go for More Information?

If these trends are interesting to you and helpful, the complete Adobe report is available for download here→

Q3 2026 Ai Sourced Traffic Insights

7.8MB ∙ PDF file

Download

Download

Also this weekend we surfaced 8 news items that are on-topic for these trends that you should read if you find these topics of interest:

What’s Next?

We’ll continue to keep a very close eye on Claude’s implementation of Agentic Commerce, stay tuned for more details as we find them!

Subscribe now