Implications of Amazon Working With OpenAI and ChatGPT

Andy Jassy released a few important details in the last earnings call that I wanted to unpack further. Here are a few pillars:

- Amazon feels that on-site agents will continue to be super-important and it's growing fast for Amazon -- and the people using it are high-intent.

- Amazon feels that LLMs are currently best for research and narrowing down options.

- Amazon will eventually partner with the public agents.

Let's start simply:

- this is very similar language to what Shopify used when it said that "we are speaking with Amazon Buy With Prime" in checkout, and the world freaked out.

So, an agreement is coming soon. It's only a matter of when.

Regarding the role of agentic in commerce. Here's my take:

agentic research and finding is an acceleration of digital influence on purchases which can happen anywhere. Considering there hasn't been agentic "commerce" before 5 minutes ago, I expect this to be the bulk of things for quite a while, despite the growth of commerce.

Also, it's likely that research-oriented users are tilting towards ChatGPT now because it's one of the best research tools ever invented, already. Which, just because someone is doing research even with "commerce intent" has nothing to do with “am I ready to buy now”.

So people should tap the brakes on conversion conversations.

Finally, Amazon is in my opinion still positioning itself to "deliver for the world". In a world where OpenAI has the best index in the world, and Amazon has the best supply chain in the world, I'm not sure how OpenAI harms Amazon in any material way.

Do we think that OpenAI understands commerce more than Amazon?

No. Do I think OpenAI will focus more on commerce than Amazon? Also, no. Ditto Walmart.

So in commerce, I do not expect OpenAI will halt the rise of marketplaces. Ultimately, OpenAI could either:

- end up fueling more purchases on existing marketplaces, or

- ultimately end up becoming a marketplace (rather than just a front-end to a merchant's checkout), in which case marketplace dominance just continues unabated.

Also, I think the "retail media is dead and agentic killed it" conversation is a little overdone. Anywhere there is a surface which displays a SKU, there will be retail media. Period. The thing will morph, but death is not something I would predict.

"But what if OpenAI disintermediates Amazon? Wouldn't this hurt their ads business?": Ultimately I do not see their ad business shrinking.

For a long time, the number of digital channels which influence both physical and digital purchases has been increasing from few to 7-10 in the past several years (so many sources). Do we think this number will shrink to one or two?

I do not. First, Amazon could take share from other players. Second, Amazon could come up with a clever way to share ad profits with OpenAI.