In this episode of The Intelligent Leader, discover how retail media is reshaping the future of commerce, advertising, and personalization. Ali Miller, VP of Ads Product at Instacart, shares how Instacart is building the ad platform of tomorrow.
In this episode of The Intelligent Leader, host Shashank Garg sits down with Ali Miller, VP of Ads Product at Instacart, to explore the rapidly transforming world of retail media. With a background at Google and YouTube, Ali brings a unique lens to how technology, data, and commerce intersect in building an impactful advertising ecosystem.
Ali shares how Instacart is expanding its retail media platform across thousands of partners, leveraging AI and ML to drive real-time personalization while maintaining consumer trust and data privacy. She discusses key strategic bets, such as their CTV campaigns with NBC and their use of clean rooms to preserve user privacy. The conversation offers deep insights into incrementality measurement, full-funnel ad strategies, and where generative AI fits in the future of retail.
This episode is a must-listen for leaders in e-commerce, advertising, media, and data science looking to understand the full scope of what's next in digital commerce.
Key Quotes:
"Creating these shoppable moments where you can inspire someone while they're watching the Paris Olympics, for example, which was when virtual concessions really came to life and helping to kind of, you know, just enrich that experience by making it not just a typical ad experience, but hey, like get your snacks to watch with the incredible game that you're about to see on your tv.
You know, it's easy to fall in love with technical solutions. It's easy to get excited about, oh, this capability that we have that we didn't have before, but. Only once it's really connected to something of value, that's where we're starting to see it come to life more."
-Ali Miller
:Using generative AI to create much more metadata, much more context and supplement, you know, your products and transactions with additional metadata that is gonna further make your AI better.
The future of personalization is less about more data and much more about smarter joints. And who you join with, how you align your purpose, and how you will continuously learn from these signals to make it better next time."
-Shashank Garg
Time Stamps:
Links:
0:00:08.2 Shashank Garg: Good morning, good afternoon, everyone, and welcome to the Intelligent Leader. This is your host, Shashank, and with me I have Ali Miller. Ali is the vice president of Axe Product at Instacart, which is a grocery technology powerhouse helping reshape how brands connect with consumers across the digital and physical retail. She leads the strategy and innovation behind Instacart's rapidly growing detailed media platform. Ali, welcome to the show. And I really looking forward to this conversation, especially given how our worlds intersect at the intersect of data, media and commerce. Beyond her work in tech and commerce, Ali is also a classically trained violinist, I understand.
0:00:50.8 Ali Miller: That is right. In fact, we have a concert this week, I'm playing on Friday.
0:00:58.4 Shashank Garg: Awesome. She performed with the Oakland Symphony. And people tell me that you're able to bring in like a unique blend of analytical rigor and creative harmony to everything you do. So welcome, welcome once again.
0:01:11.2 Ali Miller: Oh, thank you so much. And now, now that I've been so flattered, I'm extra excited to be here, but very excited to talk about everything. As you said, the intersection between our worlds and the increasingly connected world between retail media and regular digital media. So lots to discuss.
0:01:26.3 Shashank Garg: So just to set the context for the audience, Ali, our team at Infoseps has spent many years helping both large media giants sort of modernize their technology and data platforms. So moving from what used to be a very fragmented way of doing business, you know, bi channel, linear, digital, and hence everything was fragmented. Right. The audience, the inventory, the measurements, right. To like a unified, you know, way of doing things. And that's been a huge transformation still going on. And not every media for is out there. We'd be one of the lucky ones who got this opportunity to build this out. And on the retail and CVG side, working with many, many retailers now and actually also starting to integrate, helping some of them integrate their inventory, real time inventory and product data with marketplaces like yourselves like Instacart, you know, that's where all that began. Far more discoverable to millions and millions of online shoppers, not just through your marketplaces, but it has a effect on their overall sales, you know, through their own distribution channels. Right. So that's been great. But let's begin with your journey first, Ali. So just help the audience understand, you know, how does your role as VP of Axe Product, what does it entail and how that has evolved with the evolution of the company Instacart as well.
0:02:50.3 Ali Miller: Yeah, I mean that's a... Take it as a great tee up and you know, this transformation of the Industry and the rise of retail media and kind of how we're bringing more data and more context together for retailers, for brands, for consumers is very much at the heart of what I do. So I've been in the digital ad space for quite a while. I came from Google and YouTube prior to Instacart. Instacart was my first deep foray into ecommerce, into the wild world of grocery and really getting to understand what makes cpgs tick, what makes brands tick, what makes this industry tick and how can we supercharge it really by bringing this technology together to help consumers discover products that they love and then help brands sell more of those products, help retailers sell more of those products. So my core job as when I'm leading the ads product team, the wonderful ads product team at Instacart is really looking at how we can take the best of that technology and bring it to life in a way that really resonates with these CPG brands and retailers who have in many cases, many, many more decades, maybe even a century of experience than Instacart does, in this industry that is so core to many of our daily lives.
0:03:57.3 Ali Miller: And so something that I've really loved is marrying that context that I was able to build up at Google and the ads industry overall. So, you know, leading into ML, leading and increasingly into AI, of course, looking at the role of the consumer journey and how to inspire at different points throughout the traditional consumer funnel and the increasingly non linear funnel that consumers go through when they're buying a product. I love bringing that to life with the actual ability to deliver a physical product to the consumer's hands that can immediately respond to when they see that ad get into their hands in an hour or less and really close that loop in a very, very real, real world, physical way for the brands who are advertising with us.
0:04:36.9 Shashank Garg: So great. And I can relate to that evolution, Ali. And as I was talking earlier, and all of this is sort of powered by underlying technology and a lot of it is about marrying data across channels. And while technology is also a large part of it, it is also the change that you have to go through with everybody who's involved in this whole process. And we saw that firsthand at this media company, right? And you know, these are... Media companies have existed for a while and the way ads are bought and sold and the way audiences are looked at, you know, there are certain processes that have worked over the years. And you know, I must say that while building the platform was complex, but being able to do all that change management and drive much more unified ad sales experience for media companies and on the retail side for them also. Now, to be able to, you know, not just look at in store ads, which is what they were used to for many, many years, and look at these retail media networks as a coherent part of strategy is a big change.
0:05:51.1 Ali Miller: Definitely.
0:05:52.2 Shashank Garg: And I think you guys are at the very heart of driving that change.
0:05:55.8 Ali Miller: I'll just add, I think it is a huge change and I think really anchoring on that the value has been the key. I think it's really easy to fall in love with a technical solution that we're building or the really cool solution that, you know, we've uncovered from more of an internal perspective. But really leading with, to your point, what is going to resonate with brands, with retailers, with consumers, and then building from there to use the best technology platforms that we can offer to solve those problems that that seems to have, you know, really helped us to focus on the right things and hopefully build solutions that are actually resonating with the industry right now.
0:06:29.5 Shashank Garg: So let me sort of take a step back and you know, as you guys built out the platform and obviously, you know, scaling this retail media ecosystem across thousands of partners is not easy and you know, on both sides. Right. So in your mind, what has been the biggest challenge and or a few challenges and how are you guys sort of, you know, working through them?
0:06:52.6 Ali Miller: There have definitely been a lot of challenges to solve and that's what keeps this work really interesting. So, when I think back to when I joined Instacart roughly three and a half years ago, we were really at the point of looking at how does Instacart continue to become an essential part of people's lives after this wild experience that we all had? You know, living through the pandemic, seeing Instacart's extreme rise in usage, and really becoming a household name. So that was kind of around the time when the basics of the ads platform were being built out. Brands were really starting to lean in and became impossible to ignore this whole ecommerce thing and Instacart in particular. And so when I joined, what I was hearing from a lot of brands was that they were starting to grapple with this concept of, okay, I'm investing, I'm leaning in, I'm showing ads. What is it actually doing? What is the actual true business impact of showing ads? I see these attributed sales, I see this return on ad spend. What's the true incremental impact? And so that was really one of the first problems that I dove into was looking at how could we take an industry leading approach to incrementality and look at what was actually going to happen if we did the true data science blast methodology of having a holdout and exposed group and looking at the difference in total sales between them on a brand by brand basis.
0:08:10.1 Ali Miller: And so first step was to measure it. We weren't sure exactly what we were going to see in the beginning and we were actually very pleasantly surprised that our baseline incrementality was quite strong. And so what we ended up doing is working closely with a number of brands to measure that incrementality, measure it on a platform level. And we're really proud that on a platform level our sponsored products incrementality is about 15% sales lift on average and many brands see, you know, more than double that. And so we've been really happy to see that come together, continue to lean into that as a really important part of understanding how truly effective ads can be. So I'd say that was one big chunk of a problem space that I wanted to dig into. From there on out there's then, you know, what other problems can we help solve for brands? How else do we build this platform taking advantage of all of this consumer usage and attention and opportunity that we have in a way that's continuing to resonate with what brands need to actually do in order to achieve their business objectives. So from there on out, I really see that that second act of Instacart on is building out more of a full funnel solution.
0:09:13.6 Ali Miller: So looking at, on the platform, how can we not only provide, you know, great search ads, but also great recommendations as you're browsing, Introduce rich media ads and inspirational formats, shoppable display, shoppable video. We've since followed up with recipes and occasions to help actually with some of that storytelling around the product that CPGs are famously so incredible at and really industry leaders at. And so that was really the second big area focus for us. And now there's this whole question which I'm really excited about and which you hinted at at the beginning, which is around this broader ecosystem of retail media. So how do we take this first party data that Instacart has? How do we take this purchase context and ability to close the loop that Instacart hasn't opened it up kind of beyond our platform? This is of course a big trend in retail media, but from a marketplace perspective, working across, you know, our, the 800 retail partners on our marketplace, how do we really help to open that up? To help brands reach the right audience off platform, use the, you know, the catalog context that we have at Instacart to show the right product to the consumer and then actually convert back on Instacart.
0:10:24.4 Ali Miller: That's really the moment that we're in right now, looking at how we can extend that offering beyond our walls. And then similarly, as we'll also be able to discuss a little later, looking at how we can offer our technology to retailers who want to play in this ecosystem by offering our carrot ad solution, which takes what we've built as our ads platform, offers it to retailers to use on their own and operated sites, thereby extending reach and efficacy for brands, but also allowing retailers to really play in this space in a way that's also helping to simplify the buy for brands. So that's sort of the act that we're in right now is I really think of this as the moment where retail media is becoming more of a capability across the industry and not just about owned and operated advertising on a specific retail site.
0:11:08.3 Shashank Garg: And that's great that, you know, you've taken the approach to make it a, not an Instacart theme, but, you know, just really making this as a capability that both retailers and advertisers can, you know, take advantage of is great. I was actually recently reading about your, your partnership to launch those virtual concessions during Olympics. And, you know, that is amazing. And I'm assuming that you guys are doing it across the board. If you don't want to, if you could just talk a little bit about what you guys are doing there, I think that'll be useful.
0:11:41.7 Ali Miller: Yeah, and that's actually been a really fun part of this whole journey, is that we're working hand in hand with the Instacart marketing team, who is also pushing the boundaries of what it takes to create these really compelling experiences on connected TV and premium TV inventory, and creating these shoppable moments where you can inspire someone while they're watching the Paris Olympics, for example, which was when virtual concessions really came to life. And helping to kind of, you know, just enrich that experience by making it not just a typical ad experience, but hey, like, get your snacks to watch with the incredible game that you're about to see on your tv, get that delivered in your hand, you know, in an hour or less, and then opening that up to CPG partners as well, to make sure that we're really working with our partners, our media partners, our brand partners, to further what we can do in the industry. So really looking at tapping into Instacart's first party data to enrich the targeting the ability to show the, you know, the relevant item using our catalog data and then convert back on the app with a QR code or a clickable, you know, interactive shoppable experience.
0:12:45.8 Ali Miller: These are things that we're working on with Roku, with CTV, inventory on the Trade Desk, with YouTube. So there's a number of really compelling moments where again, we can take that storytelling that's so inherent to the TV experience and can really capture consumers attention in all the moments that they're not actively grocery shopping, inspire in the right moment, but then make it actionable and then by nature more measurable as well. We've been really excited to see that both in the virtual concessions work that you mentioned and then generally across the industry as well.
0:13:17.3 Shashank Garg: And we've been very fortunate to be working with both media firms and retailers. And I can, from internal standpoint, you know, I can relate to, you know, what you're saying. And when people see that lift, right. That is sort of more rewarding and it's worth all the complexity. So, you know, although you're saying this in three minutes, it takes a lot to make this, make this true. There's just too many integrations that can go wrong. There is in the measurement frameworks, you know, several things can break and you can less report. There's just too many issues, you know, and working through the trenches, I have some of our teams, you know, work through behind the scenes a lot of these things and I know how incredibly complex it is. But at the end of the day, when you make those experiences happen for your clients and the audiences, there's nothing more satisfying.
0:14:06.9 Ali Miller: Absolutely. To the consumer, it should all be completely seamless and look so obvious and so natural. But to your point, there's a lot of very serious work that goes into the data integrity, maintaining consumer privacy, making sure that we're passing only, you know, the information that we are comfortable, you know, sharing with the media partner and taking all of that incredibly seriously, going through all the grunt work and making sure the data is ready to use in this way, in a safe way. That's such an important foundation to all of this. And so it's been really exciting to see. Then it comes to life in this way that feels really natural and exciting to a consumer.
0:14:44.1 Shashank Garg: I'm going to dive into, you know, the sort of two topics now. Maybe we can go a little deeper on what enables everything that you spoke about. Right. And one is the personalized experience. At the heart of it, it's personalization. Right. And while there is a cost to personalization and the cost is that you're giving away your data and then you're trusting the media company, the retailer, the marketplaces to do what is right and what you have consented to with that data to sort of ensure what I've consented to. So if you can talk a little bit about how all this is happening in a privacy compliant way, I think that'll be very useful.
0:15:29.5 Ali Miller: Yeah. So we take, of course, privacy very, very seriously and in particular, we make sure that the actual personally identifiable information does not go anywhere it's not, if not intended to. So we're the only party that's really collecting the first party data. We have a very important trusted relationship with consumers that we need to maintain. So, you know, we utilize, you know, solutions alongside our media partners. So, you know, partners like LiveRamp, using Clean Rooms alongside our media partners, making sure that we feel really good about only ultimately exposing the aggregated, anonymized data back to the brand to make sure that we're not passing along anything that we shouldn't. So of course, at a very high level, making sure all of the, you know, consumer privacy policies and disclosures and everything are very big part of the foundations. But in actually setting up these integrations with our partners, we make sure that, you know, we are the party who continues to own the, and protect the information of the consumer and that we're only passing that aggregated anonymized data back to the brand. At the end of the day, what we've typically done when we're looking at some of these integrations that you're talking about is we create segments of audiences that really resonate with brands.
0:16:42.3 Ali Miller: So as an example, something that brands look to us to provide would be something like this consumer has browsed the category before but has not purchased my brand, or this consumer has purchased my brand before but hasn't purchased it recently, has purchased something, you know, that suggests that they might be interested, but, you know, they would be a potentially very useful incremental, new to brand user. And so we're able to create these segments of audiences that are then able to target, able to be targeted across the platforms that we're partnering with. And then on the other side, we're then able to say, oh, you actually got 10% new to brand purchases. You were able to see this particular behavior for consumers who had not purchased your product or brand in the past on Instacart and now you're able to see that aligned with the specific campaign they were actually now showing that purchase behavior. So that's been a really important part of really proving the value. It's not, you know, it's not the true incrementality measurement that we're able to do on the Instacart platform, but it really starts to add much more context around the true down funnel efficiency and efficacy of these campaigns that brands are running across multiple platforms.
0:17:51.7 Ali Miller: And we've seen that there's a lot of hunger for that. I think. You know, I very much believe in mid funnel and upper funnel metrics as well, depending on what the advertiser goal is. But being able to marry it with that down funnel context and understand, hey, this actually led to some purchases that I'm now able to understand and actually take into account that can help with a much more well rounded understanding of a true full funnel experience and campaign where you're bringing a lot of different capabilities and context together. That was previously just impossible prior to these marriages of retail media data and these platforms.
0:18:24.8 Shashank Garg: Yeah, and actually it's hard to not thank the emergence of the data breeding technology. I remember the early days, you know, three, four years ago, when this technology had just come out. Right. And people were skeptical and you know, as we delved more and more into it and the platforms got mature in their abilities to do mashup data in a clean room, you realize that personalization is no longer just about a single touch point. It's about stitching together behaviors across commerce, content, advertising and then activating it responsibly.
0:18:56.4 Ali Miller: Agreed. And I think clean rooms, you know, one of the people on my team has this running joke of like, maybe this year is the year of the clean room. Like we've been saying it for so long, but it just shows, it kind of goes back to the earlier point of, you know, it's easy to fall in love with technical solutions. It's easy to get excited about all this capability that we have that we didn't have before, but only once it's really connected to something of value. And that value could be any, you know, it includes protecting consumer privacy, protecting personally identifiable information, protecting, you know, the ability of the data to stay, you know, on the side where it was collected. In addition to opening up all of this opportunity for measurement for more, you know, meaningful targeting and personalization off platform, that's where we're starting to see it come to life more. And it does seem like we're on a road where that's going to increasingly be part of how we activate these sorts of capabilities across the industry and across media channels.
0:19:50.3 Shashank Garg: Actually, in my experience, what I'm realizing is that in addition to the tech, right, the ability to align consent, identity and attribution in a way that respects the customer's trust is key. And in our company at least, you know, at Infocepts, we joke that the future of personalization is less about more data and much more about smarter joints and who you join with, how you align your purpose and how you will continuously learn from these signals to make it better next time is where the holy grail is.
0:20:27.1 Ali Miller: Yeah, I think there's a lot to that. The other thing I would add is I, you know, this is a good moment to bring in generative AI. We do need to talk about it. And I think the thing that I find so compelling about behind the scenes, using some of these LLM based techniques, technologies and smarter context that we can add to these experiences that we're building is that you don't always need to collect more information from the user. There's so much real world context, you know, particularly in a grocery experience, what ingredient goes with what? What concepts can we create out of a... What might otherwise look like an unrelated set of items that now looks like snacks for running or you know, what to have before you go to Zoom in the morning? Like all of these sorts of things that you should require manual curation. I love the idea that we're able to take, you know, all of the sort of engagement then feedback groups that we've typically had in ML based ads development and then marry that with the sort of real world context that AI can bring in. Plus all of these joins, as you mentioned, to connect the dots across different media experiences and help the consumer feel like they actually do have a truly personalized experience across all of these different touch points that sort of knows or can anticipate what they're interested in.
0:21:41.6 Ali Miller: I feel like it's a really powerful future that again doesn't require collecting a bunch of additional PII. And it really is about bringing that real world context to life in a way that resonates with our day to day behavior.
0:21:53.7 Shashank Garg: So in the consulting world and in the data AI consulting world, we talk about AI for data and data for AI. So what you're referring to is just using AI, generative AI to create much more metadata, much more context, and supplement, you know, your products and transactions with additional metadata that is going to further make your AI better.
0:22:16.5 Ali Miller: Exactly.
0:22:17.6 Shashank Garg: Your cycle that if you get this right, you're not reducing manual effort and you don't even have to collect half the data you're collecting.
0:22:25.7 Ali Miller: Exactly. And then again, to the earlier point of like bend to the consumer, it just feels right. It doesn't, it doesn't scream at you that, you know, this is an AI based experience. It's just something that feels more relevant and feels like something you might be interested in engaging with. So I love when those pieces can all come together in a way that really just creates a better consumer experience versus screaming about one technology or another.
0:22:47.0 Shashank Garg: Absolutely. Moving on. Ali, let's talk a little bit about, in all these complexities and all the exciting experiences is measurement. And you talked about incrementality and measuring that is not easy.
0:23:00.3 Ali Miller: Yes.
0:23:01.7 Shashank Garg: And measuring that and you being able to convince the advertiser that this is correct is even harder. So anything that you think you're doing there that people can learn from, please go ahead.
0:23:16.7 Ali Miller: This has definitely been a huge focus for us. And one thing that I love about CPG is they have very high standards and a lot of industry experience about the type of measurement that actually resonates. And we've seen this through the rise of all different types of advertising. So online video, like from my YouTube days, that was definitely an evolution over time. It goes from being a hot new thing and sort of measured in whatever way is possible to something that really needs to align with standards and expectations and accreditation over time. Same with social. And then now we're seeing that with retail media. And so like I mentioned, you know, incrementality was really that first area where we went really deep on understanding how can we go beyond what the industry is doing now, borrowing from approaches that we've seen in other areas of digital ads and looking at how can we convince ourselves that we're driving true incremental value here and then starting to take that to CPGs, going deep with their analyst teams or their data science teams and making sure that they felt good about our methodology, that we were as transparent as possible and really partnering on this new capability.
0:24:21.4 Ali Miller: And we're now seeing that that's kind of going across the industry in a much more cohesive way. In addition to that, you know, providing more tools to average for advertisers to be able to understand when we're launching a new capability. So, for example, optimized bidding, we launched that, you know, several years ago now. But it's something that has been a really important part of how we drive campaign success. So instead of managing your cost per click, your mass cost per click individually you set your target return on Aspen and we can manage bids on your behalf to achieve that target return on Aspen. Very common capability in the industry, not so common in retail media at the time that we launched it. And so we built AB testing to help to prove okay, if you're running a manual campaign versus an optimized campaign, how is that doing in terms of the rise you're achieving, the total sales you're achieving? And we've kind of rinsed and repeated that playbook when we've introduced a flavor of optimized bidding to drive new to brand sales, for example, and additional goals that we're building out now. So I think that's a really big part of it is being transparent about how are specific features working for you, what are the metrics you want to monitor, for example, total sales in ARM A versus ARM B to understand if that's actually driving value.
0:25:35.5 Ali Miller: And how can we be transparent about the methodology that we're using. So it's very clear that even though this is using our own on platform data, than the risk of grading our own homework. Of course, how can we present that in a way that really feels like it can be trustworthy? So I think that's one big piece of it is looking what can convince us where can we take kind of the latest and greatest ads technology and use this to either prove efficacy of certain features or prove out total effectiveness of the platform? The other piece I'll add is accreditation. So, so as we've seen the rise of retail media, many, many retail media networks out there, we've seen that the push for standards and external accreditation has risen alongside that. And so that was something we actually, I'm very glad we anticipated that very early on because these are challenging processes. So we achieved MRC accreditation last year as the Media Ratings Council. That is a very deep and involved process of deep auditing of all of our, you know, our in, you know, spam workflows and you know, the end to end process of what happens from retrieval to serving an ad, et cetera.
0:26:42.8 Ali Miller: And we were one of the first retail media networks to achieve that accreditation. That while it's not measurement in itself, building that trust in my clicks and impressions are real. My spam reporting, invalid clicks, et cetera, are taken care of in a way that's accredited, accredited by this external body. That's also been a really important angle, particularly for some CPGs who really value this as part of any sizable investment in a platform. So that's been a really big part of it. And then finally now we're in this world of bringing more of that measurement and positive capability across the ad platforms that we're powering now for retailers in addition to beyond our walls with the off platform partnerships we talked about and then even in store with our smart caper carts that are now roaming the aisles of the grocery store and being able to also serve ads and show, you know, additional inspiration to consumers. So it's a really important part of making sure that we're building out these capabilities that resonate with brands. We're providing kind of the core day to day metrics that they understand. But we're taking it one step further by focusing on building trust and building additional proof points that these ads are actually going to drive value for their business.
0:27:55.1 Shashank Garg: That's great. And I can relate to all of it and I, you know, so for us as a firm, we've been in business for two decades now and some of our early clients were actually measurement companies. And I remember the days when, you know, all the measurement was based on surveys, physical surveys.
0:28:12.6 Ali Miller: Which have their role, to be clear. That is also really valuable. But yeah, alone that can't quite tell the story.
0:28:18.6 Shashank Garg: And that was the era then when all everything that advertisers wanted was just, you know, impressions and clicks and that's, that was the world and Google Analytics and all that. Right. But now demanding incrementality or looking at audience overlap diagnostics. So that's something you started doing now.
0:28:35.7 Ali Miller: Exactly. Yeah. We've been partnering with Circana for example to look at the incrementality over in store behavior. And we're seeing really, really promising, you know, 75% of folks in, you know, looked at in store versus online are actually buying different brands on Instacart versus in store. So that's definitely another angle. We're looking to partners in the industry to help tell and help connect the dots where maybe we can't alone. That's been a really important part of also creating value.
0:29:03.8 Shashank Garg: Simple definition of returns on ad spend, you know, total money spent versus, you know, total revenue or vice versa has become very, very complicated because now you have to fit in the incrementality. You have to fit in any sort of causal lift that you are seeing across, you know, between different experiences in an omnichannel world.
0:29:24.6 Ali Miller: Exactly.
0:29:24.9 Shashank Garg: I was studying that. You know, I think where this is going, and I'm assuming you're doing through some of your tools is to help advertisers you know, and brand managers to be able to predict, you know, what cohorts or what audiences are most likely to respond to a certain, you know, way of doing an ad. Right. And that's where you know, so from a reactive to proactive measurement and now going into, you know, a little more predictive as to what you should expect is where I see the measurements moving.
0:30:00.1 Ali Miller: Yeah, and that's very aligned with the approach too that we've been taking more on the ML and AI side where we actually do have a, we have a fairly automated platform in a lot of ways. So on our platform where we can understand and predict how consumers may be interacting with they're retailer of choice and searching and browsing, we actually rely a lot on an algorithmic approach to show the write out at the right time, which on platform in that context will almost always beat out a manual form of targeting. Off platform there's a much broader set of context and we're still learning about what resonates perfectly. So I don't think we're yet at that point where it can be fully algorithmic, but in partnerships with many of these media companies and platforms that we're working with, we're also looking to provide context to help them show the right out at the right time. Knowing the catalog information behind the brand, the brand's presence on Instacart, taking away that complexity, where is the product available, where can I get it quickly but then still being able to lean into those algorithmic approaches on those platforms. So there's a lot of that kind of collaboration and you know, pushing the industry forward, happening across all of those different touch points.
0:31:11.9 Shashank Garg: So moving on from AI algorithms to agentic AI, we have to like, when we talk about generative AI, we have to talk about agentic AI. So just curious, any early experiments there? I know the technology is still early and you know, in much more predictable, you know, workflows, agentic AI solutions are becoming more of the norm or will become faster. But this is such a rapidly evolving space. Just curious, part of any experiments there.
0:31:41.1 Ali Miller: We're definitely, I mean the Instacart team is very passionate about using the latest and greatest technology. We've been, you know, an early partner of OpenAI's and many of our, you know, early, early experiments and agentic AI is no, no exceptions. So you know, partnering early on operator and really understanding what is it that consumers want to accomplish here. How do we learn together? I think really leaning into the industry is the right approach at this point, like learning, trying new Technology, seeing how it's resonating, seeing how adoption is kind of morphing over time. And thinking back to the early days of ChatGPT, it was sort of like, okay, generative AI is always going to be about chatbots. It's going to, everyone's going to build a chat bot and they're going to use it for every single experience. And then over time you learn, okay, this is maybe really great in a support context or maybe a few examples of kind of more conversational search, but behind the scenes is where we're going to see maybe the most resonant experience on the Instacart side as an example. And I think with agentic AI, where we are so early in understanding what is it consumers will feel comfortable with, at what scale, for what purpose.
0:32:47.3 Ali Miller: And so we want to lean in and test and learn and then also look around in the industry, in the ads industry, in the consumer products industry and ecommerce as well and understand where we can all go together. Because it's quite a bold, brave new world and I can't predict exactly where it's going to go. So the best approach that we found is just to continue testing and learning together.
0:33:07.1 Shashank Garg: I can share a couple of examples there that we've been experimenting with. We built like a AI based media assistant for a retail media partner with the intent that you can autonomously analyze historical campaign data, simulate different target and other strategies and propose sort of optimal configurations for example. Essentially turning every campaign brief into a performance driven plan in minutes. Right. I mean that'd be ideal, you know, that'd be ideal. But you know, obviously it's gonna take a lot of burning for it to be anywhere close to production.
0:33:39.3 Ali Miller: Definitely.
0:33:40.4 Shashank Garg: People are starting to think about it. Gives me a lot of hope. And you know, these are complex tasks, but if you have consistent data, I think someday, you know, you will get there.
0:33:49.6 Ali Miller: I've generally found as well, you know, I'm really excited about a lot of the internal productivity opportunities that AI provides us. You know, whether it's, you know, we, everyone now like you visit for meeting notes and action items and things like that that are just, it just takes a little bit of extra effort out of your day. The other piece that seems to be particularly resonant right now is looking at how we can drive more efficiency and ease of use in advertiser facing tooling. We've been working with many emerging brands for a long time and many of these smaller brands who are looking to break through and just simply don't have the time or the people to do all of these specialized tasks of setting up, managing campaigns, monitoring that some of the larger, more established brands have. So taking the effort out of creating a landing page for their display campaigns, for example, is something that we did with AI generated landing pages. We have the catalog context, we have the context of what they're advertising in the campaign. And as long as we're able to see that it drives performance, and in fact it drives even better performance than the manually generated campaign landing pages.
0:34:50.7 Ali Miller: That's a great example of using the technology again, in a way that really furthers a specific use case and something that's of value to the advertiser. And so that's another area where we're leaning in a lot. Not quite in the agentic space, but certainly using the latest and greatest technology to create again that context in a way that a consumer can understand and engage with. That saves time for the advertiser too.
0:35:11.5 Shashank Garg: Absolutely, absolutely. One of the area that I'm starting to see a lot of traction with agentic AI is sort of assisting in sort of budget allocation. So when brands want to campaign across multiple major media networks, theoretically agents can optimize in near real time based on sort of return, return trends, then supply availability and seasonal demand shifts. Right. So can they do that? Is again an experiment that we are running and we'll find out. But regardless of what happens, you know, I think the promise of agentic AI is to ensure that they can not only help humans make some of these decisions very, very fast, and when we create them, we ensure that they are explainable, controllable, aligned, the advertiser goals and consumer trust.
0:36:00.8 Ali Miller: Exactly, yeah. Going back to the trust point, which is really the through line across all of this, like trust in measurement, trust in results, trust in the platform, and then trust in potentially an agent who's acting on your behalf.
0:36:11.7 Shashank Garg: So, Ali, just wrapping up, I know we spoke, touched upon a lot of interesting points. Just looking ahead. Right. And sort of what trends, technologies do you think are going to become more and more prominent? I think we touched upon a few of them and also anybody in this whole space, what advice, if at all, would you like to give companies?
0:36:37.0 Ali Miller: Great question. Looking ahead, there's a couple of things I would mention. One is, you know, I talked about this huge rise of retail media and the many, many retail media networks, hundreds of retail media networks that brands are now working across and trying to make decisions across every day. I believe that we're definitely at a point of simplification of that ecosystem and you know, leaning toward what are the trusted platforms, what is driving the most valuable outcomes for me? How do I manage this in a way that makes sense? This is where we're really leaning in with our carrot ad solution. So kind of creating this one stop shop where brands can buy an Instacart ads campaign that is actually able to run across 200 plus retail media networks from the partners that we work with. For example, we are working with Thrive Market that we announced last year. We have Hy-Vee coming up. And so we've been really excited about all of the momentum in this space. So that's an area where I think that that additional push on again, standards, trust, ability to drive scale in a way that's simpler to manage. I think we're very much at that moment in time, we're coming out of the rise of retail media network explosion.
0:37:44.7 Ali Miller: We're going to see a little bit of that simplification. So that's a big one. The second one is, I don't even really think we said omnichannel yet in this conversation, but omnichannel is obviously very much top of mind for particularly brands who are looking to maximize their ability to reach consumers where they are and drive value in a connected ecosystem. And so, you know, looking further ahead as we provide, you know, additional capabilities in the aisles of the grocery store with our smart capercards that are able to bring that more dynamic and personalized experience to life actually in the aisle, in addition to the digital aisle. That seems to be an area where both across Instacart and across the industry, there's a lot of interest in kind of bringing the best of retail media online kind of into the store where it makes sense, where it resonates with consumers to create this more cohesive activation and measurable opportunity for brands, but also a more connected experience for consumers. That's a big opportunity for retailers of course as well. So I think that's the second big one that I would, that I believe is really on the cusp of taking off. And there's a lot of really promising early examples of momentum there. But you know, looking out three, five years, I think we're really going to see it take off.
0:38:55.6 Shashank Garg: That's great. And I would agree to both of those. I'll just add a couple from my experience, Ali, what I'm seeing is, and I do want to call out the tech platform players who enabled the interoperability. I mean this whole thing works with the interoperability and the data clean groups, collaborations. Right. And so right from Snowflake, Databricks, AWS.
0:39:17.1 Ali Miller: Exactly.
0:39:17.8 Shashank Garg: Google Microsoft, all of them have played a part here. And you know, I hope this technology continues to get better. This sort of powers all this. And also to a lot of the retailers and rosters. Right. Who are looking to work with, you know, either building their own retail media networks or working with third party retail media networks. You know, don't just look at this as ad tech. You know, build it as a connected customer first, consumer first capability with commerce, content, data, all sort of working together, you know, working with. This is all about collaboration and partnerships.
0:39:54.7 Ali Miller: Completely agree.
0:39:56.0 Shashank Garg: It is complex work to create the platform, but if you don't have anybody on the platform, then this platform is good for nothing. Right?
0:40:01.8 Ali Miller: Exactly.
0:40:03.3 Shashank Garg: It's a very different way of doing business. But you know, it is exciting and you will see the rift.
0:40:08.9 Ali Miller: I very much agree. Yeah, it's really all about resonating with the consumer and that is actually again, kind of going back to what I love about the ecommerce space. Ads in the ecommerce space is that ads truly can be information and useful and you know, real world products that you're actually buying and using and kind of integrating into your life. And I think it's a really important thing for us all to remember when we're working in the numbers and the roas and the incrementalities like these are actually products that are improving people's lives and we're trying to make that a little bit easier for them.
0:40:41.6 Shashank Garg: Great. Thank you for the great conversation, Ali. This was super exciting, super useful. I'm sure our audiences are going to love this.
0:40:51.4 Ali Miller: Thank you so much for having me.
0:40:53.3 Shashank Garg: Thank you again to Ali for joining us on the Intelligent Leader podcast. For our listeners, if you enjoyed this episode, please rate this and review it on your favorite platform, wherever you're listening and see you again soon.