First off, the main selection "top picks" display showed a price of $229 that doesn't even exist on the offer card. Of even more interest to me is the offer card's second presented option from Samsung at $219 - that is likely a completely different monitor which has been misclassified or the price information is just wrong in the OpenAI data. I did a quick search on my own and ran into mismatched prices when linked to off-site purchasing (I didn't get the offer to complete within the chat conversation), which took me several corrective steps to get it to finally admit had the wrong prices which I only pushed because I knew it was wrong. That is likely to turn into a poor buyer experience and further validates the need to have monitoring systems ensuring that the presented information from LLM outputs matches the reality so buyers do not get an unpleasant "shock" when going through further steps to make the purchase and especially since the complaints are likely to land on the sellers since they will actually be responsive to the shopping public.
I also noted that in your video the advertised price point of $48.99 suddenly dropped to $39.99 as you were going through the checkout flow, so another place where some sort of stale or incorrect information was being used. While that was in the consumer's favor, the higher point on the offer card could have discouraged the buyer from even starting in the first place or sent them to another seller.
yes - the point of the video was just the checkout - everything you point out, EVERYTHING, is fixable by the merchant, most don't know it's all messed up -this is the whole point of ACO I've been writing about for the last yr. The industry will take time to figure it out - they are obsessed one/two levels 'up' with GEO 'mentions' - wouldn't you rather get the bottom of funnel right? We'll get there - but great observations!
We hope it is fixable by the merchant, but trying to find out how the wrong information was surfaced and how to get it corrected is going to be a much larger challenge than it has been historically with shopping systems. That will take a whole new skillset, and exactly why what you're doing is going to be important and valuable to a whole cohort of businesses that don't even know that this is a problem they are facing and should already be addressing.
First off, the main selection "top picks" display showed a price of $229 that doesn't even exist on the offer card. Of even more interest to me is the offer card's second presented option from Samsung at $219 - that is likely a completely different monitor which has been misclassified or the price information is just wrong in the OpenAI data. I did a quick search on my own and ran into mismatched prices when linked to off-site purchasing (I didn't get the offer to complete within the chat conversation), which took me several corrective steps to get it to finally admit had the wrong prices which I only pushed because I knew it was wrong. That is likely to turn into a poor buyer experience and further validates the need to have monitoring systems ensuring that the presented information from LLM outputs matches the reality so buyers do not get an unpleasant "shock" when going through further steps to make the purchase and especially since the complaints are likely to land on the sellers since they will actually be responsive to the shopping public.
I also noted that in your video the advertised price point of $48.99 suddenly dropped to $39.99 as you were going through the checkout flow, so another place where some sort of stale or incorrect information was being used. While that was in the consumer's favor, the higher point on the offer card could have discouraged the buyer from even starting in the first place or sent them to another seller.
yes - the point of the video was just the checkout - everything you point out, EVERYTHING, is fixable by the merchant, most don't know it's all messed up -this is the whole point of ACO I've been writing about for the last yr. The industry will take time to figure it out - they are obsessed one/two levels 'up' with GEO 'mentions' - wouldn't you rather get the bottom of funnel right? We'll get there - but great observations!
We hope it is fixable by the merchant, but trying to find out how the wrong information was surfaced and how to get it corrected is going to be a much larger challenge than it has been historically with shopping systems. That will take a whole new skillset, and exactly why what you're doing is going to be important and valuable to a whole cohort of businesses that don't even know that this is a problem they are facing and should already be addressing.