In this episode
Rahul Shah, CEO of CommerceIQ, argues that discoverability has replaced shelf position as the place brands win or lose in retail, because shoppers now start with questions to AI shopping assistants rather than keywords. He explains why product content and retail media ended up in separate teams with separate budgets and KPIs, why that split now costs growth, and how CommerceIQ closes the loop by piping retail media signals into product content across thousands of SKUs in near real time. Shah also covers what keeps automated content consistent with the brand, how CommerceIQ attributes a 2–12% incremental sales lift through Amazon-run A/B tests, and what a brand has to publish when the first evaluator of its product is a machine.
Key takeaways
- Shoppers now start with questions, not keywords. Shah says AI lets people ask natural questions and get specific answers back, which is why the shift toward discoverability happened in so short a time.
- AI shopping assistants on retailer sites are where the impact is today. Of the many forms of “AI shopping,” the assistant on the retailer’s own site is where CommerceIQ sees the most traction, with shoppers who start there far more likely to convert.
- The split between content and retail media is a legacy of static content. Content was maintained and optimized for compliance, retail media drove awareness and some conversion; in Shah’s view both now serve the same goal of top-line growth.
- Content can now be updated 10, 15, 20 times instead of once or twice a year. AI makes it practical to optimize content across a brand’s full breadth of SKUs and retailers repeatedly, which is what breaks down the old silos.
- Retail media signals should reach product content in near real time. When certain terms perform in retail media, CommerceIQ pipes them into content across all relevant SKUs and retailers through its AI orchestration layer, Ally, instead of waiting weeks or months.
- Brand context is the guardrail for automated content. CommerceIQ’s forward deployed engineers work with 10 to 20+ stakeholders at each brand to capture context beyond documented guidelines, and that context accumulates rather than leaving when people change jobs.
- Brands keep the final say on what gets published. Shah says brands can dial automation up or down depending on how much confidence they have built in the output.
- CommerceIQ reports a 2–12% incremental sales lift from its content agent. Shah says the lift is validated by A/B testing that Amazon runs at the brand’s request, which CommerceIQ cannot directly influence.
- For shopping agents, current content is the signal. Brands that dynamically update content for seasons and circumstances get surfaced sooner and higher by agents shopping on a person’s behalf.
Chapters
- 01:05 — When an AI assistant does the shopping
- 02:54 — Rahul Shah’s background and new role at CommerceIQ
- 04:03 — What CommerceIQ does and who it works with
- 05:15 — Why discoverability became the battleground so fast
- 07:03 — Which AI shopping surfaces are moving volume now
- 08:59 — Where uniting retail media and product content breaks
- 10:59 — Closing the loop across thousands of SKUs
- 14:16 — Keeping automated content sounding like one company
- 18:02 — Letting brands dial human control up or down
- 18:51 — Isolating the 2–12% sales lift for the CFO
- 21:07 — What a brand owes a machine evaluator
- 23:00 — One year out: the rate of change
- 24:18 — Staying agile: reducing the gap between customer and code
Why discoverability moved so fast
Shah’s explanation is that shopping behavior changed at the first step. People used to type specific keywords to start a search; now most ask questions, and AI returns more specific answers. He notes that if the technology had not been good enough, the shift would have taken much longer. He cites rough figures on AI-started search and on shopping through Alexa as evidence that the change is showing up in numbers, not only in sentiment.
Where uniting retail media and product content breaks
Asked whether data, budget, or decision rights break first when a brand tries to unite content and retail media, Shah says all three, weighted differently by brand. The separation reflects an older operating model in which content was static and maintained for compliance while retail media drove awareness. The customers he sees with breakthrough top-line growth over the last one to two years are treating the three silos as one question: what content motions and what retail media motions together maximize growth.
How the closed loop works across thousands of SKUs
In the past, a signal on one side of the silo could take weeks or months to affect the other, if it ever did, because large brands run hundreds or thousands of SKUs. Shah describes CommerceIQ taking terms that perform in retail media and applying them to product content across every relevant SKU and retailer in near real time, using seasonal examples like Halloween candy searches. The same loop is used to optimize retail media budgets for return on ad spend, coordinated through CommerceIQ’s AI orchestration layer, Ally.
What keeps automated content sounding like one company
Shah’s answer is context. CommerceIQ’s forward deployed engineers go beyond written brand guidelines and interview 10, 15, 20 or more stakeholders to understand what a brand is trying to accomplish, then train the operating model on that context. Because the context is cumulative, it is not lost when people at the brand move to other jobs. Brands still decide whether content is released, and can loosen or tighten that control as their confidence grows.
How to isolate the sales lift for a CFO
Shah puts the incremental sales lift from CommerceIQ’s content agent at 2 to 12%. He stresses that the A/B tests behind it are run by Amazon at the brand’s request, serving the same or different content to audiences, and that CommerceIQ also uses years of data and statistical modeling to account for seasonality and other day-to-day variance. His message for CMOs in budget season is that this is an independently tested number, not a marketing figure.
What a machine shopper needs from a brand
When an agent is the first evaluator, packaging and brand story carry less weight. Shah says the difference comes from publishing updated content and relevant context, staying consistent with brand values while changing content for specific times of year and circumstances. Brands that update as frequently as possible give shopping agents the signal to surface their products sooner and higher.
FAQ
What does “discoverability is the new shelf” mean? It means the deciding moment in retail has moved from physical or page placement to whether AI-mediated search and shopping assistants surface a product at all. Rahul Shah argues that shoppers now start with questions, so products that systems can’t surface never get considered.
Which AI shopping surface matters most for brands right now? According to Shah, the AI shopping assistant on a retailer’s own site is where CommerceIQ sees the most impact today. Shoppers who start there convert at much higher rates.
Why should retail media and product content be managed together? Shah says the separation is a holdover from when content was static and retail media drove awareness. Now both serve top-line growth, and signals from retail media should update content across SKUs in near real time.
How does CommerceIQ keep automated content on-brand? Its forward deployed engineers capture brand context from 10 to 20+ stakeholders and train the model on it, and that context accumulates over time. Brands keep final approval over what is published.
How is CommerceIQ’s 2–12% sales lift measured? Shah says it comes from A/B tests Amazon runs at the brand’s request, supplemented by statistical modeling for seasonality and other variance.
What do shopping agents need from brands? Shah says they need updated content and relevant context, refreshed for specific seasons and circumstances, so agents can surface products sooner and higher in results.
About Rahul Shah
Rahul Shah is CEO of CommerceIQ, the AI-native retail growth partner for brands. Rahul has spent more than two decades building and scaling technology businesses across product, engineering, and go-to-market. He was most recently Chief Product and Engineering Officer at Marqeta (NASDAQ: MQ) and held senior leadership positions at PayPal and GE. He also scaled venture-backed companies including Branch Metrics and Applause past $100 million in revenue. He holds an MBA from Berkeley Haas and a B.S. in mechanical engineering from the University of Massachusetts Amherst.
Rahul Shah on LinkedIn · CommerceIQ Names Rahul Shah CEO to Lead Next Phase of Growth
Resources
The Agile Brand podcast is brought to you by TEKsystems.
The hardest part of martech right now is telling which tools are actually earning their keep. That’s the question running through the MarTech Summit Boston agenda on October 27. We’re attending MarTech Summit Boston on October 27 as a media partner. If you want to join, the code AGILE15 saves you 15%.
Need to evaluate martech, AI, or CX platforms and don’t know where to start? The Agile Brand Guide List is now in beta. We’re building the most useful, free-to-use, and up-to-date way to assess what truly matters when evaluating platforms.
Reach your customers with Reddit. Spend $500 in ad spend, get $500 back in ad credit!
Chaser is the only Slack-native project management platform that helps teams turn messages into tracked tasks, automate follow-ups, and maintain team-wide visibility, without adopting another tool. Now integrated with Claude and other GenAI tools. Learn more at trychaser.com and use code AGILEBRAND for a 3-month free trial (normal trial is 14 days).
Start building your own apps with Replit and get $20 off.
Enjoyed the show? Tell us more and give us a rating so others can find the show.
Don’t miss a thing: get the latest episodes, sign up for our newsletter, and check out Martechipedia, the wiki for marketing technology.
The Agile Brand is produced by Missing Link—a Latina-owned strategy-driven, creatively fueled production co-op. From ideation to creation, they craft human connections through intelligent, engaging and informative content.
Transcript
Title: CommerceIQ CEO Rahul Shah on discoverability as the new shelf
[00:01:05] Greg Kihlström: Hi, I’m Greg Kihlström, host of The Agile Brand, and here’s a question for you. When an AI assistant is doing the shopping, does your product still have to be the best one or just the one the system can read, verify, and recommend? Agility used to mean reacting quickly to what the market did. Increasingly, it means operating at the tempo of the systems that decide what a shopper even sees in the first place. Today we’re going to talk about why discoverability, not the product page, has become the place brands win or lose in retail, why keeping retail media and product content in separate teams has turned from an org chart quirk into a competitive liability, and how to automate at retail speed without the brand starting to describe itself differently in a hundred places at once. To help me discuss this topic, I’d like to welcome Rahul Shah, CEO at CommerceIQ. Rahul, welcome to the show.
[00:02:41] Rahul Shah: Thank you for having me, Greg. It’s a pleasure.
[00:02:44] Greg Kihlström: Looking forward to talking about this with you here in a minute. But before we do, why don’t you give a little background on yourself and your new role at CommerceIQ?
[00:02:54] Rahul Shah: Sure. I just recently took over as CEO of CommerceIQ. I’ve been with the company six months. I joined as president and COO, and I spent the last six months talking to customers across the globe and many of our teammates to get a better appreciation for where we were at and what we wanted to do. And so I’m excited to talk about what CommerceIQ is going to do in the future. But just to give you a quick point on my background, I’ve spent the last thirty years working in and around technology, helping companies scale. My last role was as chief product and technology officer at Marqeta, a public fintech. And prior to that, I’ve spent decades at large companies like GE, Allianz, and PayPal, and smaller startups, helping them scale well past a hundred million dollars.
[00:03:46] Greg Kihlström: Love it. I know we’ve had some guests from CommerceIQ on in the past, but for those that may not be as familiar, could you tell us a little bit about CommerceIQ? What’s the company’s core focus? Who are the enterprise customers you typically partner with?
[00:04:03] Rahul Shah: Absolutely. I’d be glad to. So at the core of what CommerceIQ does is that we are the AI native retail growth partner that helps our brands not only improve their rankings in search but also end up helping them convert further down the funnel. And so we work with various different brands from across the globe. We have over twenty-two hundred brands that have done upwards of thirty billion dollars in sales that we work with. And we are very, very excited to be able to actually go through and help them influence not just top-of-funnel search, but bottom-of-funnel conversion with AI at its core.
[00:04:48] Greg Kihlström: All right, so let’s dive in here. We’re gonna talk about a few things today, but I wanna start with one of the things that I teed up in the beginning, and really, it’s this discoverability as the new shelf. So you’ve stated that discoverability is the new battleground for brands. What actually changed to make that shift happen this fast, and what are teams still measuring that no longer tells them whether they’re winning?
[00:05:15] Rahul Shah: It’s a great question. The thing that changed is exactly what you and I probably do every day. In the past, we used to go to type in specific keywords to begin our search process for whatever it is that we were buying. Now, most people are asking questions rather than typing in keywords. And the fact that AI is allowing people to be more natively asking basic questions to get more specific answers back is what has caused the massive shift in so short a time. If the technology were not perfect or were not good enough, it would’ve taken much, much longer. But now, I went to buy something yesterday, and I started my search process off with a question. And that question then led to a much more detailed view on what I should look for. So that shift is not only just thematic, but it’s proving out in numbers as well. I think it’s nearly one in seven are using some kind of AI capability to start their search process. Nearly 300 million people have used Alexa for shopping in the recent past. These numbers all speak to the tectonic shift that’s taking place in so short a time.
[00:06:33] Greg Kihlström: And of course, we say AI shopping, right? But that could mean many, many different things depending on who you are, what you’re using: a retailer’s ranking model, an assistant that helps build a basket, an agent that transacts without a person ever seeing a product page. Which of those is actually moving volume for your customers right now and which is still mostly in pilot phase?
[00:07:03] Rahul Shah: Well, I would say all are evolving. But the ones that we’re seeing have the most impact are those that exist in the AI shopping assistant that exists at retailers’ sites. And so, of course, to your point, Greg, AI shopping is getting very, very broad, and so there are multiple ways to define that. But what we’re seeing most concretely is the fact that while people may start off across multiple surfaces, where the most traction is felt is when someone goes to an AI shopping agent at a retailer and starts asking those questions. The conversion metrics themselves speak volumes about when someone starts off that way, their likelihood to convert is exponentially higher. So while many of these areas are being evolved over time, this particular piece is where we’re seeing the maximum impact, and that’s what brands are asking us every day, which is, “How do I show up better when someone goes to Alexa for shopping or any other retail shopping agent? And how is it that we can make sure we’re not only discovered in the process, but then we also convert?” And therein lies the convergence between content optimization and retail media optimization to make sure that flywheel effect does work.
[00:08:28] Greg Kihlström: Well, let’s talk a little bit about the content and retail media as well ’cause I think product content and retail media, they didn’t just end up in separate teams, they ended up with separate budgets, separate vendors, even different KPIs or definitions of success. So based on what you’re saying here, when a brand tries to unite them, where does it break first? The data, budget, decision rights, maybe all of the above?
[00:08:59] Rahul Shah: That’s a great question. I actually think it’s all of the above. And depending on the particular circumstances that each brand is at, it might have a little higher weighting with one versus the other. But for the most part, the context of it being different data sets, the context of it being different budgets or organizational structures, it was all a function of an old way of thinking and an old way of operating, where content was relatively static and was meant to be maintained and optimized for compliance purposes, and retail media was meant to be able to drive awareness and some conversion. In the new world, in the new norm that we’re talking about, they are all serving the same master associated with growing top line. And a specific example of that is when we’re able to optimize content not just once or twice a year across all of your SKUs, but you’re able to update content 10, 15, 20 times across your full breadth of SKUs across all retailers. That is what’s changing with AI, and to be able to marry the two, the old silos are beginning to break down. One thing that we have seen is at the customers that we have who have had breakthrough success over the last one to two years with top-line sales, they are beginning to break down those three different silos and look at it as one. What content motions do I need to have, and what retail media motions do I need to have to optimize my top-line growth? And our ability to do both allows them to be able to maximize their incremental sales uplift.
[00:10:37] Greg Kihlström: So closing that loop then, what does that look like in maybe a real example here? What does the system see in retail media performance, and how fast does that loop actually close for a brand like you’re mentioning, running thousands of SKUs across several retailers?
[00:10:59] Rahul Shah: Let me articulate the problem statement first, and then I’ll talk about the solution. In the past, that cycle time used to take weeks, sometimes months, or sometimes it just never happened because a lot of the large brands have hundreds if not thousands of SKUs. And so whenever you did see some behavior on one side of the silo, the ability for that to take effect on the other side of the silo, for all the reasons you talked about, Greg, would sometimes happen very slowly, if happen at all. Now, when you’re able to see that on the retail media side you’re having great success with certain specific terms, phrases, words, et cetera, we’re able to take that information and pipe that onto the content side in near real time across all of the appropriate and relevant SKUs across all retailers. And so if you do happen to see that mint happens to be something that’s very important for whatever reason, or we’re coming up on Halloween, and there’s something that’s very important for buying candy, whatever those terms are, we’re able to not only place the appropriate keyword optimization there on the retail media sides, we’re also able to pipe that onto the content side because we have that closed loop capability. And thereby elevate both, not just content optimization to maximize sales lift, but also optimizing your retail media budgets to get the maximum benefit and return on that ad spend. And our system, which we call Ally, which is an AI-based orchestration layer, allows for both to happen seamlessly very quickly.
[00:12:38] Greg Kihlström: A new landing page turns into a pile of tickets and handoffs. Framer helps your team move faster. Teams can collaborate in real time, iterate on the same page, and publish instantly. Agents help you go beyond the vibe-coded site and a first draft to bring you a production-ready site faster than ever. Framer is the AI website builder that helps creators, teams, and businesses ship production-ready sites faster than ever while getting every detail right. Prompt, inspect, edit, and publish in one place at a whole new pace. Agents and Framer work alongside teams to streamline collaboration on the same canvas, build custom code components, create and manage CMS content, optimize SEO settings, audit for issues, and ship everything all in one place. Learn how you can get more out of your site from a Framer specialist or get started building for free today at framer.com/agile for 30% off a Framer Pro annual plan. That’s framer.com/agile. Rules and restrictions may apply.
[00:14:16] Greg Kihlström: Of course, when you’re talking about hundreds or even thousands of SKUs, you’re talking about lots of content updates. When content updates are firing automatically across thousands of listings, there’s a lot of considerations, right? So there’s some risk that not only an error, but it could be more subtle, that a brand starts describing itself slightly differently in 100 or more places. What keeps automated output sounding like one company, and what does a human have to keep owning for that to be true?
[00:14:55] Rahul Shah: One thing that’s absolutely fantastic about not just AI, but the platform that we’re building as well, is the topic of context. The more you’re able to train an AI model and platform, the more it’s able to replicate more accurate results so that humans have to do less. And so when you take a look at a particular brand, one of the things our forward deployed engineers do when we deploy our capabilities at a brand is we go deep into understanding the context of that brand. Now, that’s not just in brand guidelines that are documented, but it’s in really going deep with 10, 15, 20 plus stakeholders at these brands to be able to understand truly what they’re trying to accomplish. What we then do is we take all of that context, and we train our operating model to be able to take that into account. So when content optimization does run on our platform, it takes into account the nuances that each brand wants, and that is important and critical to that brand value. And so the more we train that model, it continues to get better on top of itself. So it’s not just that you train it once. It’s the more you train it, the more it simulates what a person can do, and it starts getting better than just a person because at brands, people move on to different jobs, context is lost. With our model, that context is not lost. It’s actually cumulative so that as brands continue to evolve and they want more and more perspective provided, we take that information and we build our modeling and our context engine around that, so the recommendations on content are as close to what they want as possible. One thing I will say, Greg, is even while doing all of that, we still allow brands to have the ultimate control over whether to release that content or not because we wanna make sure that while we are an AI native company, we also wanna make sure that we give brands the control to be able to do what they feel is right. So we’re able to make it much easier for brands to optimize their content. We’re building in as much context as possible to do that as well, to do that accurately, reliably, and fast. But the final decision, we’re always wanting to work with the right leaders over at the brands to make sure that they ultimately have the say in what gets pushed.
[00:17:37] Greg Kihlström: That sounds like a good balance because, in addition to humans, just the sheer volume that you’re dealing with, right? You’ve gotta have some assistance in order to be able to do that and do it consistently. But having that human in the loop there seems beneficial as well.
[00:18:02] Rahul Shah: Especially at this point in time, as different retailers, different brands, and different customers are all learning how to adopt this technology. And so what we wanna make sure we do is we move at the pace that’s as fast as technology is evolving, but we’re also building in the flexibility in our platform to make sure that we balance that pace with the pace at which our brands feel comfortable. And different brands have different thresholds. The more they’re comfortable with it, the more they allow for things to pass through because they’ve built enough confidence in what they’ve seen and what we’ve done for them where they let that through. But as some of the brands that are learning to adopt AI or are still adopting those practices want more control, they’re able to dial up and down based on the flexibility that we’ve built in.
[00:18:51] Greg Kihlström: Well, and I think another component here too is you’ve cited anywhere between 2% to 12% sales lift from dynamically aligning content with shopping trends as well. So we’re talking about brand consistency, human in the loop, but also this dynamic alignment with what’s really happening. How does a CMO get to that number? How do you isolate that lift from all the other moving pieces? We’ve got a lot of CMOs listening here, and it’s budget season for many out there. What’s the number a CMO should be putting in front of that CFO when we’re talking about some of this stuff?
[00:19:35] Rahul Shah: Absolutely. So what we have been able to see is there is anywhere from a 2 to a 12% lift in incremental sales as a result of using our content agent, period. That number is statistically proven and independent because the A/B testing that’s done is not done by us. This is done by Amazon at the request of brands to make sure that what we are doing actually has the lift. So the most important thing anyone who’s seeing this podcast can take away, the CMOs in particular, is that that uplift is not something that’s a marketing number that we’re just saying. It’s a proven, statistically valid number that comes directly from Amazon’s A/B testing, which we don’t have an ability to directly influence other than them sending the same content or two different content to their audiences to understand which one turned out to be better. And so the fact that we actually take that into account, we do also have years of data and statistical modeling to account for other variances in day-to-day shopping. It could be seasonality, it could be other things, to make sure that what we put in front of brands is statistically proven and reliable. That’s what brands and CMOs at brands should take away, that we can deliver true uplift that’s validated by A/B testing, which is run by an independent perspective.
[00:21:07] Greg Kihlström: Love it. So the last thing I wanna talk about here is certainly we’ve touched on this a little bit before on the show, but it’s this idea of when the shopper is a machine, or at least when part of the shopping is done by a machine. As agents take on more of the buying and more of the buying process, the first evaluator of your product isn’t always a person, right? It’s a system that can’t be moved by packaging or even a compelling brand story. What does a brand need to make available to that type of evaluator that most aren’t publishing today, or at least publishing readily today?
[00:21:50] Rahul Shah: It’s what we’ve talked about for most of this call, which is the ability to actually publish updated content and relevant context makes the difference for brands to be recognized, not just by humans, but also by shopping agents that are doing work on behalf of those humans. As we’ve talked about, the ability for brands to be able to make those changes, on the one hand, you wanna stay consistent with your brand values, but on the other hand, the more you can dynamically change your content to the specific times of the year, the specific circumstances that might come up, the better off you’re gonna be. And so in the world that we’re moving towards, the ability to dynamically update that content as frequently as possible is what will make the difference for shopping agents to really catch on to the moving trends, to be able to actually have your products showcased sooner and higher up the search chain than others.
[00:22:45] Greg Kihlström: Love it. Well, Rahul, thanks so much for sharing your ideas and insights today. Got two last questions as we wrap up here. First one, if we were having this interview one year from today, what is one thing that we would definitely be talking about?
[00:23:00] Rahul Shah: It’s the rate of change. What we’re talking about is AI assistants in the shopping process. I also come from the payment side where as much as we’re talking about AI doing everything, it’s still not fully embedded in that last step of the process. And so I firmly believe that a year from now, the shopping process itself is going to be and look and feel very different. It includes AI shopping from a consumer perspective, from a retail perspective, from a brand perspective. All of those motions are going to get more technologically enabled with AI, such that a year from now, what we’re talking about on this call will likely be table stakes, and we’ll be talking about what the next year looks like because the rate of change is picking up, not slowing down. So if anything, the one constant I would say one year from now is how quickly the model that we’re even talking about on this call is changing.
[00:23:56] Greg Kihlström: Love it. Well, let’s get you back on the show in a year, and we’ll look back on this as the good old days, right? [laughs]
[00:24:04] Rahul Shah: Will do. Greg, it was a pleasure connecting with you, and I do look forward to connecting with you next time as well.
[00:24:09] Greg Kihlström: Absolutely. So a very last question for you before we wrap: what do you do to stay agile in your role, and how do you find a way to do it consistently?
[00:24:18] Rahul Shah: Two different things. Number one, I spend a lot of my time with customers. It’s how I started this call off. The more I can understand where customers are at, the better off I can really make sure that what we deliver to them is actually adding value. There are a lot of companies out there that are talking about AI as part of who they are and what they do, but I will never lose track of understanding where we can help solve customer pain points. So number one, I spend a ton of time with customers, as much time as possible, and number two, given the rate of change with technology, I’m spending an equal amount of time with our technology team, making sure that we are able to stay on top of the latest technology changes that are taking place, and innovations, so that we can connect up what I’m hearing in the field with our customers and our teams that support them with our ability to deliver that value. And the sooner we’re able to close that gap, one of the things that I talk about within CommerceIQ is I wanna reduce the gap between customer and code, and the more we can reduce that gap, it’s by applying pressure from both sides to make sure that we can reduce that gap. That’s what I spend most of my time doing.
[00:25:32] Greg Kihlström: Love it. Well, again, I’d like to thank Rahul Shah, CEO at CommerceIQ for joining the show. You can learn more about Rahul and CommerceIQ by following the links in the show notes.
