Retailgentic | Consumer Behavior & Retail Trend

In this episode, Scot digs into a question he keeps getting from readers and listeners: What’s actually going on inside Amazon’s Rufus, and what should brands do about it?

Ian brings fresh analysis from Sensor Tower’s panel-based methodology (privacy-compliant, double opt-in), walking through Holiday ’25 shopping sessions to show where Rufus shows up in the funnel, how usage spikes during peak moments, and why it appears to correlate with dramatically higher conversion. Along the way, they zoom out to the bigger shift: as more people learn to “talk to AI” (thanks to ChatGPT-style habits), conversational shopping becomes increasingly normal, and increasingly hard to “optimize” using old keyword-era tactics.

Highlights:
  • “The consumer is trained.” A year of daily conversational AI use has taught people how to prompt, so when they see Amazon Rufus, they already know how to use it.
  • Rufus sessions show a major conversion lift. In Ian’s read of the report: ~3.5× lift vs. non-Rufus sessions, with Rufus-touch sessions far more likely to end in purchase.
  • Rufus had outsized influence during Holiday ’25. A large share of purchases included Rufus interaction, even if not every session did.
  • Their team mapped ten distinct Rufus-assisted shopping paths, including the standout “Research Conversationalist” profile.
  • Correlation vs causation is real. Ian flags measurement caveats: not every “non-Rufus session” is a shopping mission; intent bias exists; so the takeaway is directional, but meaningful.
  • Brands should shift from “keyword jail” to “product truth.” Better attributes, clearer specs, stronger review signals, and real storytelling matter more in conversational shopping.
  • AEO anxiety + the brand-level rebound. Ian argues the future isn’t just “optimize for prompts”, it may reward brands with stronger reputation and social proof signals (think forums and communities).
  • AI can talk consumers out of premium. They’ve seen examples where AI steers price-sensitive users away from expensive brands, an early warning system for brand teams.
  • Retail media won’t vanish, but it will mutate. In a lower-click world, retailers will experiment heavily to preserve value, without turning the experience into ad soup.
If you’ve been treating Rufus like a curiosity, this data makes it hard to ignore: conversational commerce isn’t “coming”, it’s already shaping how high-intent shoppers decide.

Timestamps:

00:00 – The Consumer Is Now Trained
02:09 – Why This Episode Focuses on Rufus
05:49 – Ian Simpson’s Background
06:47 – Founding a Bottled Tea Startup
09:14 – Pathmatics → Sensor Tower Acquisition
10:11 – How Retail Media Intelligence Was Born
13:50 – How Sensor Tower Delivers Its Data
15:36 – Why Sensor Tower Started Tracking Agentic Commerce
21:08 – How the Rufus Analysis Was Done
22:48 – The 3.5× Conversion Lift Explained
24:18 – Amazon’s $10B Rufus Claim vs. Independent Data
28:16 – Correlation vs. Causation in Rufus Usage
31:23 – Rufus as a Research Companion
32:20 – The Cup Holder Story (Problem-Based Shopping)
39:55 – Why Rufus Usage Sticks After the Holidays
40:19 – The 10 Rufus Shopping Profiles
41:35 – The “Research Conversationalist” Funnel
45:00 – Why “AEO” Makes Ian Nervous
46:47 – Brand Matters Again in Agentic Commerce
48:47 – When AI Talks Consumers Out of Premium Products
58:21 – Retail Media’s Future in a Low-Click World
01:01:31 – Avoiding the Minority Report Ad Nightmare
01:06:21 – Instacart: The Wild West of Retail Media
01:07:15 – Where to Follow Ian

👉 Connect with Ian: https://www.linkedin.com/in/iansimpson/
👉 Learn more about Sensor Tower: https://sensortower.com

Check out the materials we discuss:
🔗 Sensor Tower Black Friday AI Trend Update 
🔗 How Amazon’s Rufus Shaped Holiday Shopping

🧠 Want to stay ahead in AI commerce? Subscribe and follow along:
📰 Subscribe to the free Substack: retailgentic.ai
📺 Watch episodes on YouTube & Subscribe for updates: https://www.youtube.com/@Retailgentic

Creators and Guests

Host
Scot Wingo
CEO and Co-founder, ReFiBuy.ai
IS
Guest
Ian Simpson

What is Retailgentic | Consumer Behavior & Retail Trend?

Retailgentic, presented by the ReFiBuy Media Network and hosted by Scot Wingo, is where Agentic Commerce meets retail innovation. We help retailers and brands prepare for the future of agent-driven shopping with actionable news, expert insights, dxis, forward-looking predictions, and practical guidance on Agentic Commerce Optimization (ACO).

Speaker 1:

The consumer is now trained. I think ChatGPT Rufus came out in 2024. It was here for the last Black Friday. Nobody cared. Nobody talked about it.

Speaker 1:

But I think after a year of ChatGPT's meteoric growth, and in general, everyone that I know has now a year's worth of chat history at ChatGPT, whether they're professional or not, knows how to ask questions, how to shape, knows how to communicate with the AI. So now if you go to Amazon and Rufus is there, people I think know better how to talk to Rufus because it's the same construct as what, as people would use for Chateaubiti. And so the consumer is just trained at this point.

Speaker 2:

Welcome to the Retailgentic Podcast, where we explore the intersection of retail, ecommerce, and AI agents which we call Retailgentic. Our mission is simple. We are shining a bright light on the new Agenic AI shopping megatrend which will disrupt the trillion dollar digital retail market. The stakes are high. Google, OpenAI, Microsoft, Perplexity, Amazon, Apple, Meta, and Anthropic are all fighting to be the consumer AI shopping agent winner.

Speaker 2:

There are gonna be some tough decisions for online brands and retailers to make. What happens to checkout? How should payments work? Could you continue to invest in retail media networks? What does the world look like when over half of your traffic is not from humans but from AI agents?

Speaker 2:

These existential questions are the tip of the iceberg. On this podcast, we're talking to industry experts, analysts, and influencers to get answers to these questions, identify trends, and best practices. This is Retailgentic, your podcast partner in navigating our AI shopping agent future. And now here's your host, Wingo.

Speaker 3:

Welcome to this episode of Retailgentic. One common question I get from Substack readers and our podcast audience is, hey, Scot. Your content on ChatGPT, Gemini, and other Agentic commerce engine is awesome, but can you tell us more about what's going on on Rufus, Amazon's Agentic commerce offering? You're the Amazon dude. Come on.

Speaker 3:

Get with it. Well, your message has been heard, and good news. For those of you that were craving more Rufus content, we are delivering with this episode. I'm excited to have Ian Simpson, who is the senior vice president of innovation and strategy at Sensor Tower. We have scoured, through all the data available out there, and Sensor Tower has some pretty cool data on what's going on inside of Rufus that I think you're really gonna enjoy.

Speaker 3:

In this conversation, we first go over Ian's background, his role at Sensor Tower, how he got there through an acquisition, and then we dive super deep into the data. First off, looking at where it comes from so you have a good understanding of that. And then we look at the data from Holiday twenty five that they have analyzed. When we go in the data, I think you'll find there's some really good actual items in there for brands and some interesting insights. I was particularly fond of this 10 funnel profile, kind of framework that Ian has come up with as they looked at the various, Rufus funnels.

Speaker 3:

We only go through one of those because for for time, this could have been a two hour podcast if we went through all of them, but I do, we will be distributing the reports, along with this so you can go look at all the other profiles and see what you wanna do. I do have a note. If you're listening to the podcast, on this episode, we are gonna go through a bunch of data, and it may help you to have a visual to go along with it. I always do my best to paint a picture with words for those of you that are listening. But maybe afterwards, you'll go to the substack where we will have both of the reports available for you to download and flip through and follow along with with the discussion.

Speaker 3:

I hope this helps scratch that itch to get more Rufus data. And we're... Now that we have a really good relationship with Ian, this would be a great one for you to like, follow him. And, if we if we help them realize there's a big demand and, we as an audience really like this data, we'll get more of it. So help me support, that effort.

Speaker 3:

Thanks for all the feedback. I hope you enjoy this one. Enjoy this episode with Ian Simpson. Ian Simpson, SVP, innovation and strategy at Sensor Tower. Welcome to the Retailgentic podcast.

Speaker 1:

Thanks, Scot. Happy to be here.

Speaker 3:

You and I have been trying to record this, but we've been so dang busy. We haven't had a chance. I had NRF. You had CES, and then you ended up doing a day trip to kinda near NRF. It's been been crazy.

Speaker 1:

It's been crazy.

Speaker 3:

The life of Agenctic rock stars, what can I say?

Speaker 1:

I suppose so. I actually just point my finger at the conference organizers. I don't think it's on us.

Speaker 3:

Yeah. It's going get busier.

Speaker 1:

Whoever plans to make NRF on the weekend before... Right after CES.

Speaker 3:

Yep. It's been that way since 1911, I think. So I don't think they're going to change. Let's start. You and I have chatted quite a bit, but you're new to the audience.

Speaker 3:

This is your first time on the show, hopefully one of many.

Speaker 1:

Thank you.

Speaker 3:

You know I am a super data nerd. You are too. So, that's why I'm really excited to have you on. You have data. I have questions.

Speaker 3:

It's gonna be fun. But let's back up a little bit. Before we dig into Sensor Tower and your role there, what... Give us the Ian Simpson backstory.

Speaker 1:

The Ian Simpson backstory is actually like... It starts out having nothing to do with tech. I started my career in politics a long time ago. Came out of undergrad, caring a lot about international relations. Worked for the Israeli embassy in Washington, DC.

Speaker 1:

Got very into international politics. Went abroad, lived abroad. Came back, got my MBA. And then finally, sort of in the last decade, really got into the world of marketing, CPG, and then into tech. So it's sort of been this meandering path.

Speaker 1:

The first real foray into it that I had was when I started my own bottled iced tea company called Teavolution back in the aughts. We were launching this bottled iced tea with the concept that we would donate a quarter, $0.25 from every bottle that we sold to charity and ended up getting distribution across the Mid Atlantic. We were in Whole Foods. But ultimately, like a lot of startups, it didn't work out. And so from there, I was able to jump into the wonderful company, which owns Palm Wonderful, owns Wonderful Pistachios.

Speaker 1:

I started there in the consulting world. I had this stint at IBM as a consultant for a long time. Once I was working there as an internal consultant on all the brands, ended up becoming senior brand manager for Palm Wonderful. And so that was really the first time that I had worked in the realm of big business CPG, building relationships, not just from an advertising perspective, but as a brand manager, also building all the retail relationships and understanding how our wonderful company sales team operates. And so, when I left that, I decided my next big, big jump needed to sort of get out of the product space and into the tech space.

Speaker 1:

And because I'd been working so much with Nielsen data and just other data that came in about how our products were selling and all that kind of stuff. I took an interview with Pathmatics in Santa Monica toward the middle of twenty seventeen. And essentially, Pathmatics was building the Nielsen data of digital, having digital advertising data across display advertising, Facebook advertising. At the time, it stopped there. But I came on to run the consulting arm of Pathmatics.

Speaker 1:

We ended up building a ton of custom tagging projects for different clients. As the company grew, so did the team. I went from a team of one to a team of five. Then we had some acquisitions, and now we were acquired by Sensor Tower in 2021. And Sensor Tower is essentially what PathMatics was but for the mobile app world.

Speaker 1:

So where PathMatics sees every ad on the internet, Sensor Tower saw every app in everyone's phone, knows what people are doing in those apps, how often they're downloaded, monthly, daily, weekly active users. And so the synergy between those two groups seemed pretty obvious to the owners of the two companies. And so we merged, and I kept on in my role of running the professional services team. But very quickly saw that what we were doing with clients from a customization perspective was actually informing us what the market wants. And it turned into an opportunity to launch new products based on what clients needed.

Speaker 1:

And so, the first big opportunity for that came with Retail Media. We had clients for several years who were asking us to custom tag ads that were coming out of retailers for the specific products that were in those retail ads. We did that enough times to think, Oh, there's a trend here. So we turned it into a product called co marketing intelligence because it was one year before retail media sort of took off as a concept. And so we were building this mass market co marketing intelligence platform.

Speaker 1:

It turned out it was really just retail media in disguise. And so, we were first to market with a retail media intelligence product. And so, we've been able to capitalize on that success ever since. And so, based on that and based on a whole bunch of other things that happened in that company, my team got rebranded as the innovation team, we work and do consultative and custom projects for clients with all of our with the access to all of the data that we have, with the goal of identifying new and unique opportunities with that data that we can then turn into products over time.

Speaker 3:

And we have a lot of people that follow the pod that are into retail media. And so the problem is you run an ad, which is just like a digital surface, you don't you don't know what product is behind that ad. So so that you're basically helping extract that information so you could tie it from the ad being seen to the product being sold. Was that kind of the idea?

Speaker 1:

Yeah, I think it goes beyond that. The easy way to describe it is if you see a Facebook ad that comes from a Walmart handle and that Facebook ad from a Walmart handle shows a picture of Gatorade and says, Come to Walmart and pick up this Gatorade, then that's a Retail Media ad. That's an off-site Facebook ad pushed through Walmart Connect for Gatorade. You do that a couple million times, and you get a view of essentially the entire off-site universe of advertising across all the major retailers. So, we are tracking ads like that in all social channels Facebook, Instagram, TikTok, etcetera plus connected television, video, YouTube, and on-site display.

Speaker 1:

And so every product that's in any one of those ads goes into our platform, and we're tracking that for 40 different retailers now. So all the big ones Amazon, Target, Walmart, Best Buy. But we've also got a number of players in different specific verticals. So we have Home Depot and Lowe's. We have Sephora and Ulta.

Speaker 1:

We have Chewy and PetSmart, and then a smattering of others. So, we've really got a lock on the full view of what's happening in the retail media space. And right now, we are developing, for the first time, a retail search component to it, which will round out the full product view. We'll know who's buying advertising in off-site and on-site and in search across all the major retailers.

Speaker 3:

Is there like a software as a service interface that retailers and CPGs come in to run queries against this data? Or is it a big data feed you provide people? Where where does Who is the customer in this? I understand the data and how do you deliver the data.

Speaker 1:

Yeah, so most of our clients have access to our dashboard. We have four major products that we deliver on. We have a product on digital advertising, a product on mobile apps, a product on gaming. So we have intelligence on what's happening in Xbox, PlayStation, etcetera. And we have a website product where we have visits and all the kind of vital stats on websites around the world.

Speaker 1:

So, any one of our clients has access to our platform, and then depending on the package that they get, they have access to different cuts of those different products. And Retail Media sits under our digital advertising product. So if you are a subscriber to digital advertising, you can add on Retail Media to have access to all of that data. And the cool thing about it is that it's designed in a way to show you the sort of omnichannel view of what's happening in the market. So, if Pepsi is a company that you are interested in seeing, you'll be able to see how Pepsi is advertising across all of their owned and operated digital channels, across social, and then across retail media all in one view.

Speaker 1:

So you get the full campaign for any one of their products at any given time.

Speaker 3:

We're gonna dig into... You guys have a bunch of Adjunctic Commerce data. Let... Let's start there. What what lead is this Retail Media Network kind of piece?

Speaker 3:

So you kinda started digital advertising, and then you have RMN. Is that what led you into this agentic commerce kind of like starting to get curious and saying, I wonder if we could add some value here? Or was there a client that came to you and said, I desperately need to learn about what's going on? Like, how did you get here?

Speaker 1:

Yeah, so think let me back up one second and tell you about our methodology. We have a massive panel of people all around the world who've essentially given us permission to look over their shoulder as they're using their phones or their computer or their television, and we can sort of pick up on the signals that they're getting through the mechanism of just their own natural navigation. It's all very PII compliant. I want to make sure I say that. Double opt in GDPR.

Speaker 1:

We're very, very proud of our privacy policies and the way that we protect our consumers. But we have all this data flowing in, so we see the ads that they see, we see the channels that they're on, we see the apps that they see. And so all of that data is funneled into those four major products that I mentioned earlier, but we are actually picking up on signals from all over the internet that we're not necessarily productizing. And that's where my team comes in now. We are working with clients to capture whatever a client might want out of that data, even if it's data that's not productized.

Speaker 1:

So, about eighteen months ago, we started getting requests for what's happening in the world of AI in general. Who's using ChatGPT? Who's using Gemini? What kind of trends are we seeing in terms of moving away from standard search into those AI channels? And so even back a year ago, we were starting to pick up on a lot of the shifts that were happening.

Speaker 1:

I think 2025 has clearly been an inflection year for AI. People are using it now in ways that a year ago, they definitely were not. The phrase agenetic is a phrase I learned in March 2025. Like, I remember the specific moment that I learned it. And then, since then, it's the only thing anyone talks about.

Speaker 1:

I learned about MCP this year. Now people are kind of over MCP and moving into ACP and now UCP. So the world is moving really, really fast. And what my team did was sit with our leadership and sit with our engineering teams and say, We may not have products for this data yet, but we need to make sure we're collecting it. And so we've been collecting as much AI data as possible for the better part of a year and a half.

Speaker 1:

Now, in our platform, we are building AI tools for clients, starting with some of the data that you've seen, the referral data from AI into other websites, citation data. Someone asks a question of ChatGPT and ChatGPT cites a link, we pick up the link. So we're able to get citations and referrals and then look at the traffic as those referrals go out. So in our platform today, we have started to report on how often are LLMs referring people into Amazon or Walmart or Best Buy or any website that you want. But in the raw data, my team has had a field day digging through and trying to come up with new and interesting ways to view the world.

Speaker 1:

So, the last thing I'll say is, with the lead up to Black Friday, it was pretty clear that this was gonna be the first Black Friday where AI had a role to play. Whether you call it agentic AI, whether you call it AI, whether you just call it better search, AI has had a role to play in Black Friday like never before. And by next year, I can't even predict what kind of a role AI is going to have. I mean, you said it, I think, in your last podcast, you had to change your script almost every day during November because every company was announcing new products in real time during Black Friday. But I expect a lot of that's going to be settled by this time next year.

Speaker 1:

So not only is it important for us to track what happened in Black Friday, but also to hold onto it so we could do the year over year comparison.

Speaker 3:

Yeah. And what I wanna do on the podcast is kind of two two chunks. So you've put out some two two reports that are really awesome. There's what I'm gonna call the January Rufus report and then the mid December kinda Cyber Week report. And, we're gonna start with Rufus, then we'll go into Cyber Week.

Speaker 3:

I realize for those of you that listen, there's gonna be some visuals. We'll try to explain them a bit here as we go through. They are available in all the show notes in your audio podcast listener. The the Substack will have them, etcetera. And then I always put them on LinkedIn too.

Speaker 3:

So wherever it's easier for you to go find this stuff, then then it'll be there.

Speaker 1:

And I will happily shamelessly plug my LinkedIn. I'm always looking for, you know, comments, people having conversations there too. So that's

Speaker 3:

good. Throw Ian a follow-up too while while you're there. Okay. So, so I've bought up one, and and you kinda have a summary thing here, It's called How Amazon's Roof is Shaped to the Holiday Shopping. Walk us through some of the insights that you discovered as you were digging into this data.

Speaker 1:

Yeah. So what we did was we wanted to go as far beyond the vital stats as we could. We, through the panel, we see all of our panelists, when they engage with Amazon, we're picking up all the signals about what pages they're hitting, when they're using search, when they're using Rufus, when they get to a checkout page. We obviously ignore the specific products they're looking at. That's their business, not ours.

Speaker 1:

But really trying to understand sort of the full flow and funnel of how they're shopping. And it is very easy to say, like, X number of sessions included Rufus. And that was some of the early data that we were putting out, where we saw a massive lift just in terms of shopping sessions that included Rufus versus not over the Black Friday period. But in our later analysis, we actually took a sample of more than 100,000 different shopping sessions and analyzed them from the moment they came into Amazon to the moment that they left, whether they left on their own or with a product, and looked at all of the times where they engaged with Rufus along the way and kind of tried to surmise based on those patterns what benefit Rufus had. So that's sort of the foundation of the whole report.

Speaker 1:

You can see clearly some major summary stats. Right here, we saw that sessions had a conversion lift of three and a half times non Rufus sessions, with about 36% of all sessions ending a sale versus 10 or 11% that didn't. Of all the shopping sessions over Black Friday, Rufus contributed to about 40% of those. But interestingly, two thirds of all the sessions that actually ended in a purchase had Rufus involved. So an outsized influence on the overall shopping experience.

Speaker 1:

And so based on all of that, today, we've identified 10 unique funnels. I'll emphasize this is very early data, so we wanted to get this out there. We might refine things. We probably will refine things. We are evolving as long as the AI is also evolving, but we wanted to get this out for the community of people who care a lot about what's going on here to be able to be part of that conversation and to shed light on what's really happening.

Speaker 1:

I think, Scot, critically, right before Thanksgiving, Amazon themselves came out with some stats about Rufus where they said that Rufus had contributed to an annualized $10,000,000,000 worth of incremental revenue. And that was using a seven day attribution window. So what Amazon was doing was saying, anytime a consumer touched Rufus, if they bought something within seven days, they were giving credit to Rufus. And that felt to me, frankly, a little farfetched. And I think a lot of people kind of had to squint sideways to see if Amazon was telling them the whole story.

Speaker 1:

And so for us, what we did is we used a one day attribution window, so it was severely shrinking from seven days to one day. And indeed, the data looks very good for Rufus. And we can talk more about all that, but I'm not here to promote Amazon, but I am here to say that from what we saw,

Speaker 3:

At there's a 30,000

Speaker 1:

foot level,

Speaker 3:

I find this super helpful because, as you know, Amazon has built this thing that's just like a... Not only is it a black box, but it's it's a... It's built of titanium. Right? It is an impenetrable black box.

Speaker 3:

They don't want any bots to get in there. And we talked to all these brands, and they're asking, what do we do? Okay. It's great. You you know, Chat and all that.

Speaker 3:

What do we do about Rufus? And I think just showing that this is important and and some of the things that can happen here if you get it right are are important. And I I know this drives you crazy, but I I like to make sure I understand the data by asking fundamental questions. So, and you and I have had a lot of email back and forth on this. And so so so, like, let's talk about...

Speaker 3:

So you talk about a 3.5 x conversion lift. So, basically, you know, whatever the starting point is, let's say it was 2.2%, that's without Rufus. So there's this set of people that were... You know, touched Rufus in the day, and there's a set that didn't. And let's say, just to give it a real number, let's say the ones that weren't involved with Rufus converted at 2%, and that must mean if I multiply that out, then the others got enough of a lift that they were at 7%.

Speaker 3:

Now I realize that's not the absolute number, but I'm just like... I I like to tie it to absolute so I could just make sure I'm understanding what you're saying.

Speaker 1:

Yep. Yep. You're saying it exactly right. You're saying it exactly right.

Speaker 3:

Sometimes there's a causality here. Like, is that because the people that used Rufus were more in market? Or do you think that there's... You think that's a really clean AB test?

Speaker 1:

I don't. If I'm being honest, and I've said this in my in my posts, there's a few a few caveats that I think are really important. The first is a non Rufus session today could include things that don't have to do with shopping. And right now, we don't know. So, a non Rufus session could be a return, could be someone checking their account.

Speaker 1:

So, there's some variability built in versus a Rufus session, which we believe Rufus is only focused on shopping. Although, I do think you can look at past purchases and things like that. There, of course, is a question of, is the shopper that uses Rufus more likely to buy because their intent is higher. So there's a chicken and the egg question. There's a causality versus...

Speaker 1:

What's the phrase? I lost it. Causality versus correlation. So one, two, three. Of course, there's a causation versus correlation question.

Speaker 1:

We tried to isolate that out as much as we could by looking at the funnels themselves and seeing that as you map out different funnels and what people are actually doing, you can start to infer where Rufus jumped in and supported the sale. So should you take it with a grain of salt? Of course. I think any data you get on any topic in 2026, you should take with a grain of salt. Whether it's about AI or retail media or whatever, I think there's a lot of pontificating everywhere.

Speaker 1:

We're trying as hard as we can to shed some real light on what is actually happening wherever possible.

Speaker 3:

Yeah. I think the message to brands, though, is if there's if there's a high intent consumer, they're gonna use Rufus. We don't know. We don't know why or the causality or the correlation, but most likely, your product is going to go through Rufus, and you need to really focus in on that versus having twenty years of optimizing the search experience. We now have a new discovery research buying experience that is very heavily touched.

Speaker 3:

That's directionally true.

Speaker 1:

What I would say is the consumer is now trained. Think ChatGPT... Came out in 2024. It was here for the last Black Friday. Nobody cared.

Speaker 1:

Nobody talked about it.

Speaker 3:

It was hidden. It was mobile only, and you had to like... It was not clear how to... Yeah.

Speaker 1:

But I think after a year of ChatGPT's meteoric growth, and in general, everyone that I know, everyone that you know, probably, has now a year's worth of chat history at ChatGPT, whether they're professional or not, knows how to ask questions, knows how shape, knows how to communicate with the AI. So now, if you go to Amazon and Rufus is there, people I think know better how to talk to Rufus because it's the same construct as people would use for Chateappity. And so, the consumer is just trained at this point. And I think Amazon is there to take full advantage of that training. Now, how a brand can optimize for that today, I don't think I know.

Speaker 1:

I think that there is some advertising in Rufus. I don't know how it impacts the results.

Speaker 3:

That's my job. We'll save that for future

Speaker 1:

Fair enough. I also think that, in general, the experience of Rufus as all AI is a lot harder to game for a brand than your typical search engine optimization or any other kind of like...

Speaker 3:

Yeah. And I think you want to game it. I think you want to make sure it has all the product information. Because if people are using it... One of the...

Speaker 3:

Spoiler alert. One of one of the the profiles or funnels as you call them you picked out is the researchologist or you have a cool name for it. I can't remember what it is. But, you know, the the researcher. And that's how I've been using Rufus.

Speaker 3:

It's like a really good... Like, I narrowed down, you know, a small appliance to three choices. Rufus has this feature. It's like a compare products or something. I put them all in there, and it was, like, the best experience ever because I was having a conversation with Rufus about these three products, and it had so much information about them.

Speaker 3:

I could finally get an answer. Where I've I've been in the store looking at the boxes and stuff, and I'm like, I can't figure it out. I've looked at the three websites and can't figure it out. And it was it was, like, by far the best experience. But if it didn't have such good product data, which is more frequently the case, this one that it happened to, that's where brands and re...

Speaker 3:

You know, brands can have a pretty big impact. It's just making sure... Because we've gotten so many years of optimizing for keywords, a lot of times, we're not giving enough product attributes for Rufus to have enough meat on the bones to to have those those more research y conversations.

Speaker 1:

Yeah. I mean, I can give you one of my own experiences. My my cup holder in my car is built for a wide cup, and I never have one. I just have a normal cup and it always is rattling around. So I went to Amazon to see if I could get a cup holder reducer, something that was just...

Speaker 1:

And so I typed in to the search car cup holder reducer, and I got a bunch of garbage that had nothing to do with what I wanted. So I went to Rufus and I, instead of doing a search, I just said, Here's the problem I'm trying to solve for. And then I didn't even use keywords. I said, I have these cups. They fit a normal cupholder.

Speaker 1:

My car is a Kia EV9. It doesn't have a normal cup holder. It has a big cup holder. What can I get in order to shrink it? And it gave me some of the same wrong answers at first.

Speaker 1:

But when I said, Well, those aren't what I'm looking for, it came back and it gave me these neoprene stickies that I could put into the cup holder to buffer it and shrink it in, which isn't necessarily made for the purpose of the thing that I wanted, but it ended up being the thing that I needed. And so I was able to figure out my problem based on this conversation.

Speaker 3:

In a world of e commerce where it's keywords, you're trying to play battleship to figure out the keyword for what you're trying to buy. Right. And then in a world where you can just describe it, it can make all those connections for you. Yeah. But if you didn't have the product data, it can do that.

Speaker 1:

I think over time, the concept is it's not just going to know you. It's going to know your family. It's going to know who you're buying for. You know, I have Just to test it, I gave it a style question dealing with a problem that my son was dealing with that had nothing to do with shopping. And it came back with a ChatGPT style answer to that problem.

Speaker 1:

It had nothing to do with shopping, but I'm just saying it has all that same foundational data in it. And so I think as it learns the personalities of the people that you're shopping for, if you're willing to tell it, then Rufus will probably give you advice based on who those people are and what it's been able to learn from you.

Speaker 3:

Yeah. I didn't see it in this report, but do you guys have data on, does the Rufus usage map desktop mobile, the the normal Amazon desktop mobile percentages?

Speaker 1:

This data is all Android data.

Speaker 3:

Oh, okay.

Speaker 1:

Gotcha. We we get we get signals from Android. IPhone is very hard to get panel signals from. We get Android signals. We also have desktop panel, and we get desktop signals.

Speaker 1:

Work still has to be done to make them compatible. So in terms of being able to get this data out, it was more important for me to just analyze the Android data than to try to merge the data also with a whole separate set of desktop data to try to have a more complete picture.

Speaker 3:

That's good to know. And then, you know, another thing is it's really interesting because over the arc of time you've been doing this, getting into Rufus was tricky. It used to be mobile only and a little thing at the bottom. But now across the whole surface area, there's like it's hard to not get into Rufus. You know?

Speaker 3:

So below the images, it'll frequently say, do you wanna ask Rufus? If you get into the reviews, sometimes it kind of seems to be on and off and very dependent, but I've definitely seen a use case where it's like, do you want Rufus to summarize these reviews? Do you know which of those entry points is kind of like how are people getting into Rufus?

Speaker 1:

Yeah, so we will. It is on our roadmap to have that.

Speaker 3:

Put that in a when you do that, we want to have you back to understand that.

Speaker 1:

Yeah, I want to be back. Mean, what we did with this one was look more at the signals of the funnel, right? So if someone is coming from the homepage and going right to Rufus, chances are pretty good they're using it for search. If they're on a product page and they're going into Rufus, chances are they're digging deeper into the characteristics of that product. And so that be can focused on the reviews.

Speaker 1:

It could be also around the tech specs and things like that that people might ask. You don't have to dig through 17 scrolls to get to your answer about what the product is all about because RUPES just gives it to you.

Speaker 3:

Yeah. And then for those of you following along on audio, we're on slide two of the Rufus report. And here, we have two two kind of time trended line charts with some good data points, and it goes from 10/01/2025 to 01/04/2026. There's a dark green line that's kind of flat and rattling around an index of a 100. And then there is a green one that once it gets into peak shopping time, the kind of the Cyber Week becomes Mount Rufus.

Speaker 3:

So so it it surges up, and then kind of... It comes back down, but you can tell that the the delta before holiday was, you know, maybe 10%, and then post holiday is more like 20, 25%, which... So there's a lot to pick away from there. I have some insights that I think happened, but I want to... Let's hear yours first.

Speaker 3:

What's this chart telling you, Ian?

Speaker 1:

Well, first of all, it tells me that there is a steady state of shopping that's happening on Amazon, whether it's Black Friday or not. But around this time period, nearly all new shopping that came in had something to do with Rufus because the dark line here is essentially flat, and the green one spikes. Now, this is indexed to 100. The overall number of sessions, the dark line is actually higher than the green line. You and I have talked about that in our emails.

Speaker 1:

But from an incremental perspective, all new shopping came in with Rufus. Now, the other thing you can see here, this is weekly data. It's, well, it is Sunday data delivered only on Sunday. So there's like a, what's missing in this chart, you see that little mini spike between ten point five and ten point one two. On ten point five, there's a spike.

Speaker 1:

But actually, on ten point seven and ten point eight, there was a prime day. And if we had mapped out all of the daily data, you'd see another spike right in there. What you're seeing, though, is there is a bump over there. But but but clearly around Black Friday...

Speaker 3:

Would that spike be as big as Mount Rufus, or it would be the Little Foot... Rufus Foothills?

Speaker 1:

It would be like Hill Rufus.

Speaker 3:

Yeah. Gotcha.

Speaker 1:

And in fact, you have... If you look at the... I don't if you wanna call it now, but if you look at the mid December report, we have the daily data on that slide, and so you can see the difference.

Speaker 3:

Yeah. Let's go... We'll put a pin in that and get back to it. The other thing that I... When I saw this, and I think I asked you a question of this also indicate...

Speaker 3:

Because of this increased... It looks like it's sticky, right? So I get it that during the holiday spike, it was up 90%. A lot of people were using it. But then, like, off your index, 20% are now using Rufus.

Speaker 3:

And it's almost like they tried it and liked it. Now, it's become pretty sticky.

Speaker 1:

I think every major period that drives people to their computer gives them more opportunities to practice, and it's gonna end up sticking more and more.

Speaker 3:

And then here, for listeners, I'm on slide eight. What what Ian and his team at Sensor Tower have done is they've they've identified these kinda 10 funnels or profiles. We won't go through all of them, but, a, I love some of the name... Names of them. There's, like, the early decider, the cart reconsider.

Speaker 3:

So that's kind of an interesting one. And then there's the research conversationalist, and it kinda sticks out on the chart because it's the biggest circle. And, also, you know, these charts are always designed, so something in the upper right is kinda like the the place to be. Right? So this one, it has the highest share of sessions and the highest conversion rate.

Speaker 3:

So, you know, this is... If we were if we were a marketer, the way I read this is that's my number one target zone. That's the target of my bull's eye. And, yes, I want some of these other folks, but but, you know, if you win that one, you've got some some good mojo going, and then you could... Maybe then you go work on the search assistant and the product validator and and that kind of thing.

Speaker 3:

Is that a... Is that how you would take... Tackle it? Or?

Speaker 1:

It is, although I don't know how to game that. And I know you didn't like the word game before, but I'll still use it. Research conversationalists, the size of the bubble is the duration of their session. People in that research conversationalist averaged one hundred and fifty minutes, I think. We have that data on the next slide.

Speaker 1:

Yeah. So they were in Amazon for a long time. Not this last, sorry, it's this one. There, you can see the duration there, one hundred and fifty minutes. And they viewed more than 30 products in this example.

Speaker 1:

So the research conversationalist is going in with a concept in mind and is willing to have a long discussion with to get what they want. You see here, average, 13 asks of Rufus, 13 turns. If that's on average, there's plenty of examples with a lot more. And yes, they convert 54% of the time, but what that means is they are dedicated to finding something. And so if you own a particular product, I will have to say, I don't know...

Speaker 1:

And again, Scot, this is your job. I don't know how you ensure that your product is in this person's funnel and is the thing that they land on as wanting to buy, right? So, the chart makes it look like an appealing place to be. I think the reality is that the consumer is very complex. The consumer is deliberately trying to narrow in on the one thing that they care about, and they're willing to spend time and have multiple back and forth conversations with the AI bot and dig in before they pull out their credit card.

Speaker 1:

And so, I think the lesson for the brand would be around, have respect for your consumer. Know that they're not just jumping in willy nilly, but they're actually spending a fair amount of time trying to find the right product. And the more you can do to give them the information that they need to make a solid and informed decision and to support Rufus in giving them that decision, I think that increases your chances of being the product that gets landed on last in this profile.

Speaker 3:

I think what's exciting about this, I talk to a lot of people that sell on Amazon. I probably talk to 3,000 of them.

Speaker 1:

More than me, for sure.

Speaker 3:

Everyone assumes it's a game of keywords and price. That hasn't come into this conversation at all. Right? So so that's what gets me excited is we've been optimizing. This changes what we're optimizing for.

Speaker 3:

We were in kind of a gamification mode, and it was like gaming the the the the dials of ads, price, FBA, and all that jazz. And all that still matters, but this person obviously isn't price driven because they're spending... They're not spending a hundred and fifty minutes looking for the best price. They're they're Ian in Los Angeles, anonymous Ian, you know, who's looking to solve his urgent cup holder problem and willing to spend two hours so his coffee doesn't spill on his lap when he takes a left turn in his Kia.

Speaker 1:

Yeah. That's it. And Exactly.

Speaker 3:

You know, you would have paid, you know, you would have paid 5... You'd be happy to pay 5, but you would probably would have paid $25 to solve that problem because, you know, all it takes is one spilled coffee on your white suit pants and your... You know, that's now now a $200 problem or whatever.

Speaker 1:

You're making a lot of assumptions about my style choices.

Speaker 3:

I've got you into Miami Vice, but in Los Angeles, kind of a... Yep.

Speaker 1:

I don't really have the opportunity to say this often, but I think when I hear people talking about how they're going to handle agentic commerce, using phrases like AEO makes me nervous. I don't think that the that the that the consumer experience of the future is truly going to be, I don't care, even though they're building it, I don't believe that the consumer experience of the future is you go to ChatGPT or you go to Rufus and you say, I want to buy a new bicycle. Tell me about the bicycle options I have. And then it just brings up these product cards. And so it's on the bicycle manufacturers to then optimize for that prompt.

Speaker 1:

I think that prompt is probably the rarest of prompts. I think what's actually happening is that people trust their chat GPT more than they trust their therapist, more than they trust their friends, and more than they trust their spouses. And they will go in and have a ten, 20, 30 turn conversation about a lot of different things. And over the course of that conversation, ChatGPT may recognize that there's a product opportunity and then serve them up a product ad or serve them up a link to a product. But I don't think it's as straightforward as I want to be there every time someone does a product search within ChatGPT.

Speaker 1:

I just don't think that's how it's going to work. And so I think, you know, I have this theory that we've been on a race to the bottom in the last ten years in terms of the funnel, where so many marketers are going away from top of funnel marketing and true brand marketing and getting into the world of conversion marketing. And I'm starting to wonder if the solve for Agentic is back at the brand level. Because if people love your brand and they talk about it on Reddit and they're blogging about it and posting about it, that's when ChatGPT is going to see the social proof that it's actually a good brand. So, I think you can try to buy as many keywords as possible and be in there at the moment of purchase, but I think you win in the future by building a good brand.

Speaker 3:

I think that's right. The other way I frame it is, you know, think back to when you had your tea company. You you came up with that because you had an idea, and you saw a hole in the market, and you're very passionate about it. You could probably have talked endlessly about your tea product. Right?

Speaker 1:

I actually have... I'll give you a little... That's my

Speaker 3:

Oh, tea. There you

Speaker 1:

go. I... This is from our trade show.

Speaker 3:

Twist sip drink with purpose. See? You've got a give back kind of a Tom's kind of element to that product that's... So so what happened is we had this world where all these brand creators and marketers are storytellers. We we wanna tell those stories, and then we're in keyword jail where we have to boil it down to, tea, give back, and then, you know, what's the price?

Speaker 3:

Yep.

Speaker 1:

That's

Speaker 3:

right. This is the... This is not gaming the system. This is going back to storytelling. Make sure...

Speaker 3:

Because of that, who's gonna spend time, like, really, really crafting the listing on Amazon and making sure and making sure people are relieving good reviews, sending their reviews to Chatuchu Tea. You need to do all this stuff to help Chatuchu Tea understand your product so that as it's having these multi turn conversations, it knows to bring it up and and, you know, oh, you care about social causes, and you love tea? Here's, like, a... There's kind of what could be a possible solution for you. So...

Speaker 3:

Yeah. So I think it's actually gonna be this golden age for marketers. I'm I'm I'm cautiously optimistic that we're entering a really interesting phase.

Speaker 1:

There's another side to it, Scot. I mean, in our research, we saw examples of ChatGPT talking down products, too. There was an example of someone asking ChatGPT, I'm debating between these two sunglasses. One of them is this premium, I won't say the brand, but one of them is this premium brand that's $500 and one of them is still premium, but $100 asking ChatGPT if it's worth it to spend the extra money. And ChatGPT comes back and says, No, you're just buying the brand.

Speaker 1:

If you want to buy the brand, go for it. But materially, you're not getting substantively something more. And I bet you that that company would love to know that Chatchi BT is guiding consumers away from their brand almost more than they would want to know how to make sure that Chatchi BT is promoting their brand. I think now I also don't know who that consumer was, but if ChatGeeBT knows that that consumer is price conscious, ChatGeeBT is probably gonna guide them away from expensive brands no matter what. Whereas if that was a luxury purchaser, someone with a bunch of disposable income, the answer might've been completely in reverse.

Speaker 1:

So that whole AEO thing is very complicated today in a new... In a completely new way.

Speaker 3:

Yeah. If you're gonna use it, use it as a directional data tool. Do not use it to create content. I think that's that's gonna be no bueno, and that's not gonna be good. A lot of them are getting into this content creation.

Speaker 3:

It's like, oh, people aren't choosing your brand. Click here to generate some content. So then they go generate a listicle and then publish it somewhere. And and I just think that, like, self publishing review data is, like, really bad.

Speaker 1:

It's not great. I will say, we are working on it is going to take us a long time to figure this out, but we are working on how to actually inform consumers or how to actually inform businesses and brands what really matters in the AI space. And we're doing it. It's going to take a while because we're doing it very deliberately to be incredibly helpful. But also, I think it's really important, I've said this before, we care a lot about privacy, and we are not trying to make a quick buck.

Speaker 1:

There are a lot of folks out there that are frankly breaking a lot of privacy rules in order to get as much data about chatbot usage as possible. And we are... We've completely pumped the brakes on it in order to make sure that we're privacy compliant in in every way possible.

Speaker 3:

Yeah. I've seen some of the AEO guys on podcasts, and and I saw one of the hosts dig into the panel, and and they diverted away from that, which is like a little

Speaker 1:

Little red flag?

Speaker 3:

Yeah. You you offer up your... Even when you write reports, like it then. You... You're very clear in where this data is coming from, and I I like the I like the transparency that that you guys have.

Speaker 1:

We're very transparent, and we stand behind it. We talk to clients all the time who tell us that we have the most thorough approach, not only from our methodology, but also from a privacy point of view. And there are clients who, like, their legal departments won't let us they wouldn't sign with us if their legal departments don't sign off on our privacy. So, we're really, really solid on that front.

Speaker 3:

Yeah. I thought this was a good one. So so for all the folks listening, I'm now into the the mid December report, and I thought this was good. So now now we're gonna pivot from talking about Rufus. We'll we'll get back to Amazon in a second because we had a pin in that looking at the the Prime Day.

Speaker 3:

But on this one, we're looking at, you know, the, basically, citation traffic, to your point. So tell us kind of what stood out to you in this data.

Speaker 1:

Your partner, Jason, gave this speech at NRF a few days ago. And he asked the crowd, Where are you in the prediction model world of, Do you believe that we're all about to become AgenTic users? Or it's a whole bunch of much ado about nothing? And if I were in the audience, I probably would tapped somewhere in the middle because as much press and as much attention as being focused on what's happening inside of LLMs and how they refer traffic and result in purchase, the absolute numbers right now and the percentages are very low. We are at the earliest, earliest days.

Speaker 1:

And what you can see in these two charts is an encapsulation of that. You've got, in the month of November, ChatGPT referred 9,000,000 people to Walmart, but that made up 1% of all of Walmart's referrals or all of Walmart's traffic. So there's a good 70% of people who just went to walmart.com because they typed it into their browser or opened it up on the app. There's another 1015% that came from search. There's another 5% that came from email marketing.

Speaker 1:

I'm making these numbers up. I don't remember them exactly. But in terms of actual LLMs, 1%. And this is all. This is not just ChatGPT.

Speaker 1:

This is the sum of all LLM traffic that went to Walmart. And you'll notice Amazon is near the bottom there. That's because Amazon walled themselves off from the LLMs. You can see where they did that in this chart back in June. There's a big drop in referrals.

Speaker 1:

So, the question is, are consumers actually using LLMs from a search perspective or not? And again, I'll refer back to what Jason said. He had a slide showing that LLM searches for products are only about 2% or something like that in his speech, but that's

Speaker 3:

still Yeah, that's a HTTP number.

Speaker 1:

Right. I remember we've all looked at that report. You know, that from an absolute value number, that's still 60,000,000 people a month searching. And I believe that's true. I just think that people are using LLMs to manage nearly every aspect of their lives, and shopping still is just a piece of that.

Speaker 1:

And so, I think you can see here that the data shows that it is still a very small portion. Even these spikes that are happening in 2025, you can see the percentages going from basically nothing up to the full 1% by November. You have to remember that usage also spiked like that across this same time period. So, are we seeing a change in behavior, or are we just seeing an obvious shift as people start using LLMs for more personal things? And I don't know.

Speaker 1:

I think that, just like a lot of us have said who grew up in the .com era, you could not have predicted where we would have gotten with social media in 1997. We are all sitting here looking into a future that is so impossible to predict that when it's 2045, we will be laughing at every one of these conversations and just how far off base they were and how the society has shifted in ways that we could never predict.

Speaker 3:

The bull case is this is the beginning of a logarithmic trend, right? So obviously, OpenAI believes that Anthropic believes they're spending as if this is going to continue. They see signals on their side that, you know, yeah.

Speaker 1:

I think there's no doubt that the investment's being made to arrive there. And I think that that's appropriate, frankly. I really do. But I think we'll see if consumers go along for that ride or if they take us in another direction. Here, you can see that the hill from October 8.

Speaker 3:

Yeah. For listeners, I'm on slide two. It's got kind of a three card slide, and the big card on the left is green, and it shows the daily Rufus activity with that that kind of prime day spike. Cool. So there's tons more.

Speaker 3:

I don't wanna spend all our time going through slides. So just five five more minutes. I know I know you gotta cut out here. The, okay. So let's take all the knowledge Ian has in his head of what it looks like today and project forward a little bit.

Speaker 3:

Like, you you also live in the retail media network world in this world. You know, one of the... I know the retail media network people, a lot of them are the biggest naysayers with Genentech Commerce, and I sense that's because they're worried about this. In a world without clicks, what happens to retail media networks? What do you think?

Speaker 3:

Do you have any tea leaves on that? I'm asking the tea guy if he has tea leaves.

Speaker 1:

Fair point. I don't know about tea leaves. I can tell you what I think, what my observations are. I think retail media came of age at a time when everybody thought that cookies were going away. And the solve for a cookie less future was retailers who had audiences that you could actually know who they were, and then get the closed loop kind of reporting.

Speaker 1:

With the cookie going away, retail media is just media, right? Like, I know advertisers are comparing, you know, they're not comparing retail media players against other retail media players. Their advertisers are comparing retail media against media, right? Like if they have a dollar to spend, they're going to put it where they have the most efficacy, they can, unless there's some contractual reason, right? If you're selling at Walmart, you might be forced to do some advertising at Walmart.

Speaker 1:

So I think as this world grows, I think that the retailers are doing everything they can to be part of it, But their motivations are hopefully such that they are thinking about every way that an advertiser can advertise and trying to create the most value out of their assets that they can to create a place where it makes sense for an advertiser to advertise. So, I'm not the kind of person to prognosticate on what direction it goes in, but I can tell you that it's very, very bifurcated. We all know that. There are a million places that advertisers can put their dollars, and it is very hard for advertisers, as humans, to manage all of them. And retail media definitely made a play for those budgets at a time when it was very advantageous to do so.

Speaker 1:

And today, it is somewhat less advantageous to do so. So the question is whether or not they can actually like, Now that they've built these media companies and they have some of this expertise built in, do they have enough wherewithal to be able to pivot with the times and stay on top of it? And some will and some won't. But there's going be a lot of experimentation, which I think is interesting. And so, we'll have a lot to talk about over the next couple of years, which is great.

Speaker 1:

But I don't think the consumer cares one bit about any of this, right? Everything that we're talking about is inside baseball when it comes to the consumer doesn't think in their lives about the funnel. They don't think about consideration. They don't think about conversion. They don't think about awareness.

Speaker 1:

They're just doing what they do. They don't know that everybody else is trying to manipulate them into one of these buckets one way or another. And I have a feeling that if the LMs play their cards right, I think they're going to try to slide into a place where it is not so obvious to the consumer. None of us want the experience of five sponsored pop ups while we're in the middle of chatting. And I think it remains to be seen if we're moving in a direction where it's seamless or my doomsday scenario is like Minority Report, where you walk down the street and you just get pummeled with advertising all day long.

Speaker 1:

Think everybody wants to try to avoid that if we can.

Speaker 3:

Yeah. I feel like we've already kind of been there with the, I think Amazon's way jumped the shark on the monetization of their SERP and their homepage. It's just, like, really hard to find anything, which I think is one of the reasons that's driving Rufus usage, which leads to my next question. One of my predictions for this year is there's gonna be a... I know this is gonna happen.

Speaker 3:

I just don't know when. I think it'll be this year because I'm I'm kind of... I'm seeing everything accelerate so fast. But I think Amazon will flip, and they'll they'll basically make Rufus the default experience. The search will...

Speaker 3:

It may be like Rufus powered or something, but the part that'll be hard about that is they will have to, to solve this ad piece or else they'll for they'll have to basically say goodbye to like $60,000,000,000 or some decent chunk of $60,000,000,000 in revenue.

Speaker 1:

I have a feeling they're not going to say goodbye to $60,000,000,000 worth

Speaker 3:

of revenue. Yeah, that's a good guess. But yeah, they'll have to they'll have to find out how to fill that hole. But, you know, the... It is hard to argue.

Speaker 3:

It is filled some. You could get rid of some advertising by having people in Rufus if it is converting at 3.5 much. Right? Because there's dollars that that's spilling out. It may not be the 10,000,000,000 downstream for the seven day window, but, you know, that...

Speaker 3:

That's a material number for for, you know, what is Amazon? Like, 400,000,000,000 worth of GMV flowing through there. There... You... That's how you make it up.

Speaker 3:

One way to make it up is in transaction dollars. Do you think do think Amazon will eventually flip that and that will become the default experience?

Speaker 1:

I mean, I'm certain of it, but Amazon's UI may no longer be the UI that we are familiar with. Right? And so I give Amazon credit, I think everybody does, for being an incredibly innovative company. I don't think that they're going to screw this one up. I think they're going to experiment a lot.

Speaker 1:

They have a captive audience of Amazon buyers. I don't see Amazon going away anytime soon. So, unless you believe that some up and comer, even someone like TMU, is going to somehow unseat Amazon. I will say, when I talk to people, the bigger issue is just the plethora of irrelevant products that are on Amazon. The fact that Amazon Some people still believe that Amazon is a store and not a marketplace, but it's a marketplace.

Speaker 1:

And so, there are so many knockoff products and so many random products. I have a feeling that they're going to figure out how to make the legitimate products stand out in ways that the knockoff ones don't. And so, from that point of view, that would be a positive change in the experience for the You probably have a much stronger opinion than I do. I do not have a background in Amazon shopping. I'm not like a Kiri masters working on Amazon for the last fifteen years.

Speaker 1:

I just know it as a consumer, and then as someone who's observed it in the data right now. I do think that I think we all know, whether you use it or not, that Facebook has jumped the shark on that. There's way too much advertising in Facebook, and everybody feels that, whether they continue to go back or not. I think Facebook knows it. If you look at the app, they have created a whole new wall called Friends because your actual Facebook wall is all ads and groups.

Speaker 1:

If you want to see what your friends are doing, you have to click the Friends button to get away from it. And we see it in the data that something like, if an ad campaign gets three or four different impressions per consumer on TikTok or on Pinterest, it gets 20 to 30 on Facebook because Facebook is just pummeling you with ads. And how many people have they lost as a result of that? I think a lot. I know that Amazon doesn't want that to happen.

Speaker 1:

And neither does Best Buy, neither does Walmart. So I think they're all treading as lightly as they can while still doing a lot. I'll tell you who is getting a ton of ads is Instacart. Instacart has if you ever go to Instacart, it's like 60% sponsored results, and our data bears that out also. So Instacart is kind of the Wild West of advertising.

Speaker 1:

If you want to do a bunch of experimentation and see, if you want to try to sell wine to cheese lovers, Instacart's the place to do it because they allow conquesting in a way that Target and Walmart do not.

Speaker 3:

Cool. Well, Ian and could talk to you for another hour, but I want to be respectful of your time and

Speaker 1:

I hope you and everybody enjoyed it. I'm trying... I am actually trying to make it as interesting as I can.

Speaker 3:

This is good. You're talking to our people. We love data. We love insights from data because it's kinda how we can squint and see the future. So we really appreciate you bringing this data to us.

Speaker 3:

Look forward to what else you published this year. If people wanna follow you, where should they go?

Speaker 1:

They wanna follow me. They can just find me on LinkedIn.

Speaker 3:

Yeah. Throw it in the LinkedIn search, Ian Simpson Sensor Tower, and it will pop you there.

Speaker 1:

But Sensor Tower is a great company that's doubled in size, you know, every every year for the last couple years. So if you really want to know what we're doing, come be a part of one of our products because we've got amazing data that helps a lot of people.

Speaker 3:

Yep. And then there's a Sensor Tower blog that's interesting. And then, you know, a couple... There's also a resource center where you can come and find a bunch of other content as well. Of course.

Speaker 3:

Yeah. And we we appreciate you taking time to put out all of this content. It's very helpful to us, and we'd love to have you back on certainly by next holiday, but hopefully sooner. If you have some interesting things, let me know, and we'll bring you on and update everyone on what you're seeing out there across the landscape of agentic commerce, retail media, and all the things.

Speaker 1:

I'd love to. Absolutely. Thanks a lot, Scot. I enjoyed it.

Speaker 2:

Follow, rate, and share to stay up to date with Retailgenic, your guide to the future of AI shopping agents.