when

you have 10 salespeople, you have 10 people effectively doing the same thing and they have territories or countless or whatever, and they're all doing the same thing.

Whereas in marketing, it's unique is that you have 10 people and they're doing 10 different things.

Um, and so the question that it ultimately is, is how important are each of those things? And is there a different way to do that? So as an example, we do, um, You know, we had an events person.

Well, we're not doing a ton of events.

Right.

And we can be smarter about events and do sort of more repeatability.

If you had to pick one or two levers that actually would move that number, what are they?

But let's start with your background and then we'll kind of pick up the elastic path background and then those obviously are going to meet up and then we'll go forward.

Sounds good.

Yeah, you know, so it's funny.

I've had maybe, I guess, a non-traditional background.

I studied journalism and film.

I lived in Indonesia making film and television commercials a long time ago.

Then I was a photojournalist.

People, BY, they bring their own or how does all that world work? It's

So the question right now is that how does a business get into the radar of the potential customers? So go ahead and educate our listeners about how are AI buyer agents changing B2B discovery, and what does that mean for structuring product data?

Yeah.

You know, one of the things that we're seeing is how prospective buyers find information is changing, right? And we're seeing the shift from sort of traditional keyword search and search engines into, um, LLMs and AI answer engines.

Um, and so it's incumbent on product companies to deliver information in ways that are consumed by the LLM, so that way they can make recommendations, right? And there's lots of solutions for what people call AEO or GEO, um- And that's important.

But what's interesting is, you know, most of the constraints we think about in terms of product information in an e-commerce setting are defined by the constraints of a product detail page, right? It needs to have a few images.

It needs to have tech specs or sizing guides or, you know, you know, related products, but it has to be, you know, typically it has to be consumable by a human.

Well, in the world of sort of AI agents and AI discovery, th-that constraint of consumability goes away, right? And the best way the AI can make recommendations for your products is when they have a wealth of information.

Yeah.

Yeah

... we don't have to make that trade-off.

Um, and so, you know, one of the things that we're excited about is that we can offer effectively unlimited attributes around your products, right? Because to make an LLM discovery process work well for you as a merchant, you know, you in effect need like an 80-page PDF worth of information about your product, so that way you can train the LLM when to recommend it, right? And it's FAQs and it's detailed specs and it's all this stuff that's well beyond what can be grokked on a product detail page, right? Like a human's not gonna read all that information, but an i- an LLM can process it in seconds.

Um, but most commerce systems don't let you sort of store an unlimited amount of product information around that product.

So we s- we have that in our, in our product experiences, and we have this capability we call commerce extensions, which is effectively like a database as a service.

So you can store any information that you want in there about products or what have you, all accessible via API.

So now you're, you know, you're delivering the, the, the meat that the LLMs need to recommend your products more often, right? So this really becomes the game, is how do I flesh out the available information about my product and get my product data models in clean and, and well structured so that I can, um, facilitate and foster these experiences and find, you know, get discovered more often by people looking for the things that I do well.

Mm