The Founder and CEO of Ibotta. With that, let's talk about the business.
Great. Thanks for having me.
After a beaten raise for Q, talk about how the business has evolved over the last year. What have you guys done differently? Talk about just improving overall operating performance.
Yeah. I think there are a few different things that we've been focused on most recently. I'd say number one is just taking a hard look at how we can execute our core business more effectively. That has meant bringing in new leadership, bringing in a new chief revenue officer, improving certain functions like business-to-business marketing, adding, you know, different kinds of sellers from a background that's more of a digital media background and reorganizing so that we're organized by vertical rather than by geography so we can understand our clients' business more deeply and provide proactive solutions more effectively.
I think that transition was disruptive for sure, now we finished that reorganization in the third quarter of last year and began to see some of the benefits of that midway through the fourth quarter. Just, you know, being, having more continuity between our sales reps, being in the room more often, providing proactive solutions around things like the SNAP program, paid dividends in terms of just our core business, recovering relative to the previous quarters. That's thing one. I think the second thing is that we really incorporated into our core business a lot of the key learnings that we had had from talking to clients over the course of the first half of the year. For instance, we took a hard look at the way that we were approaching pricing.
We took a hard look at, you know, how we were defining the goals of a campaign, promotional campaign to make that clearer upfront, making sure that we presented a metric of profitability that was more credible, so focusing on incremental sales only and focusing on the cost per incremental dollar. We built all that into our core product. Then on top of that, we had these kind of exciting next-generation capabilities, which we call LiveLift, which are related to being able to project the profitability before and during a campaign even more regularly, and then building in the opportunity to optimize those promotional parameters to solve for whatever that target outcome is. That we began piloting last year.
You know, we had more of those LiveLift pilots in the fourth quarter than in the first, second, and third quarter combined. We exceeded our own expectations in terms of the performance of LiveLift from a revenue standpoint, from a re-up standpoint, got tremendous feedback. The last thing, Andrew, is that we brought on third-party measurement, which was a really big innovation for the industry. The ability to have a third party like Circana or ABCS Insights actually produce their own, separate, you know, independent report that is called a sales lift report, which is akin to what you could get if you spent a certain amount on Google or on Facebook, but in the promotions world hadn't been made available.
The combination of better execution, improvements to our core product, some sizzle around this emerging new LiveLift product, and then the third-party measurement, particularly, kind of helped us, you know, push the business in a more constructive direction.
Let's again, bigger picture question here. Help me understand just incrementality for coupons, right? Like, this is a question that constantly swirls around the business and you guys are now better addressing. Walk me through the pressure points of what CPG companies have historically thought and kind of like the evolution of that thinking as it relates to promotions.
All right. Let's start with Grandma Boone, okay? Grandma Boone got her Sunday paper, there were a bunch of coupons in it. She went, got the scissors, she looked for the products that she likes to buy. When she found something that she loved to buy, she cut it out, put it in her purse, and took it to the store and used it. Well, that was popular with the store because the presence of these coupons was causing her to come in more often and buy more things, they encouraged these brands to continue to put these coupons in the newspaper.
The brands were uneasy about it because at the end of the day, they knew that a lot of the time Grandma Boone was gonna buy that product whether she found a coupon or not, and all you really did was just decrease the price she paid as a subsidy, which was unnecessary. The conventional wisdom that arose was you do this insofar as the retailers like it, and they'll give you something that you value, like primary placement of your product in the aisle and end cap, that kind of thing, but it's not something you want as much as possible of. You wanna do kind of what you need to do.
Along came the world of digital coupons and that changed things a little bit in the sense that, you know, you could target offers more to people who you wanted to put them in front of. Fundamentally, there were still no attempts made to show that those load-to-card coupons were profitable because they largely, these companies, our predecessors, went to one retailer and said, "I'd like to power your digital coupon program, and I'd like you to force all these brands to participate in it, and let's split the proceeds of it accordingly." That is not the approach Ibotta has ever taken, right? We came out with an app that was a direct-to-consumer thing. It wasn't done via the retailer, and it had to stand on its own merits as a sensible thing to invest in.
From the beginning, we've cared more about the performance of these campaigns. Since our IPO, we've really sharpened our focus on that and realized that what the market really wants is something that is just as statistically rigorous and persuasive as the analysis that Circana would do for digital media like any other, you know, Google, Facebook, whoever else is getting a lift study. That meant we needed to create a consensus around a metric to measure profitability, and we needed to push back on this idea that it's a binary thing, right? Is a coupon a bad or a good tactic? Is it profitable or unprofitable? The answer is depends, right? I can describe a coupon right now, buy five, get $0.25.
It's probably very profitable, but not a lot of people are gonna take you up on that, so the volume isn't gonna be that great. I can describe one that's gonna drive a ton of volume, you know, buy one, get $5. Gosh, but it's probably not very profitable. How you find the middle ground, depending on where you wanna strike the balance between volume and efficiency, exploring that and then having a more, kind of rolling ability to measure profitability. Instead of finding out 60 days after a campaign, or God forbid, six or 12 months later in a media mix model, actually knowing that, much more, helpful to know that while the campaign is happening multiple times during the campaign so you can adjust and course correct.
Then you strap onto that the power of machine learning, and that's a powerful vision. So, we've come to a point where we've built a consensus. We're building a consensus around the idea that promotions can be profitable, and as long as you're doing them intelligently, you wanna do as much of that as you can, subject to whatever constraint or outcome, you know, you've defined. That's a bit of a broader paradigm shift because it is not only challenging conventional wisdom that, you know, coupons are unprofitable, but it is opening the door to a level of kind of rule-based resource allocation that has never existed in this industry.
Talk about the use of third parties in terms of measurement, right? Specifically for Circana, right? Like what were the lessons that you learned in terms of incorporating this more over the last year, and what's that unlock been?
Yeah, we knew we heard consistent feedback that it would be better if they could turn to an independent source and that they would use their data, they would use their statisticians, their statistical methods and produce the format in their, or a report in their own format for a number of reasons. I mean, most obviously, they don't have a dog in the fight economically. Beyond that, they can benchmark, right? They can say, "This is the level of incremental lift that you're seeing relative to a control group that wasn't exposed to this promotion. Here's how that benchmarks to the level of incremental lift you would see, the median level of lift on your digital media," right?
Now, a promotion is much stronger medicine than digital media 'cause you're giving somebody something. You're, you're inviting them to click it, click on it and take action because they've got a benefit there. In fact, we did see that there was a dramatically higher increase in lift than the average digital media tactic. Because they have all that data, because they're doing the same thing for YouTube and The Trade Desk and what have you, that becomes another valuable aspect of it.
I think as much as anything inside the building of a CPG organization, being able to say that we are willing to send the data and have somebody else grade that itself is just such a strong signal of confidence that what we're doing is valuable and it is, it is at, you know, if anything, we're being conservative in how we're measuring it. I think that goes a long way to creating a sense that there's really something different going on here.
Connect that now to LiveLift, right? You've kinda talked about it. Flush out that discussion. Just help us understand kind of how it's incremental in terms of what you guys were offering before.
Yeah. What we now do in our core non-LiveLift campaigns is, in most cases, if you run a campaign with us, at the end of that campaign, you're gonna be told how many incremental sales we think you drove and the cost per incremental dollar that we think you paid. Then we can all evaluate that, and we can adjust the next campaign and so forth. That's great, right? You can buy a third-party lift study or we'll buy one for you if you spend enough money. That's all great. The thing that LiveLift does is it takes it into the kind of AI future of the industry by saying, "What if we could give you a projection, a range of incremental sales we think you're gonna hit up front?
Every couple of weeks, you know, we can check in and update you on where you're trending in terms of your cost per incremental dollar or your incremental sales. Let's say your target is $0.35 and you're coming in at $0.20, well, you're leaving a lot of profitable revenue on the table. You wanna sweeten that offer and get it closer to $0.35, so your volume goes up. Let's say you're coming in at $0.55. Well, you wanna dampen that offer because you're coming in too hot on that side. That's an example of the way you can be much more nimble and agile in how you're allocating dollars. And then there's the fact that the AI part of it, right?
If you imagine a world where you don't have a consensus metric, it's very hard to grade a promotion as successful or unsuccessful. Unless you can create a success metric that is embraced, you can't have machine learning optimizing for a metric, right? The reason why machine learning can create the greatest chess player in the history of humanity, it's artificial intelligence, is that we can all observe the outcome of a chess match, right? If you have an outcome that everyone agrees, okay, our goal was $0.30 CPID, and this is at 40, that failed. This is at 28, it succeeded. You, you tell the machine all the different parameters that it can control. How many do you have to buy? How often do we show you the offer? What's the value of the offer?
You can try different permutations until you solve for the target outcome. What you're really doing by defining outcomes of profitability is you are setting the table for a level of persistent optimization that's gonna really unlock efficiencies for these companies in the coming years. You know, we're just on the very front end of that. We've been using artificial intelligence since, you know, for a decade. As we think about optimization being done more and more by models over time, that becomes a good reason to start to invest in learning how LiveLift works, because that's gonna become one of the primary applications of AI in the CPG industry.
You talked about optimizing or setting a objective at the beginning of campaigns on the last call, that was, like, one of the key changes you guys. It did. Help me understand the evolution of that, though, right? Like, what did that look like before versus, like, what we are today? Is that just directly then related to LiveLift as we think about trying to then solve for that outcome?
It's broader than that. Let's go back a few years. You would agree to do a promotion, and then the promotion would end, and 30-ish days later you would send a recap deck. Now, all throughout the promotion they could see the performance metrics. They could see them every single day. They could see their redemptions, the budget consumed. That was true. The actual kind of profitability was a wrap-up analysis where you got a ROAS claim in a deck. Okay?
That was, I think up to a certain level of investment, people will tolerate that. You start to ask them to invest beyond that level, and they expect it to resemble the other things they invest that kind of money in, you know, these bigger digital marketing platforms that I mentioned. What happened with the goal setting was that we would go to talk to clients after the campaign, and we wouldn't agree with them on whether the campaign had been successful.
Sometimes what would happen is they would decide they wanted to maximize volume, by the time we went to evaluate the results and noticed that it wasn't that profitable, they would object to how unprofitable it was. It was a bit of a moving target. Well, wait a minute, you wanted maximum volume, we gave you that. We need to write this down. We need to write down what are the goals so that we're not debating that later on. We need to get to those results in your hands sooner, right? Automating that helps with that.
That's separate from LiveLift in the sense that even today without LiveLift, you can now define the goals up front very clearly, you can get the readout in the form of incremental sales and cost per incremental dollar at the end of the campaign, and we're not having that same issue where, you know, we're debating whether or not we achieved the mark, right? If we missed the mark, everyone knows that and agrees with that, and conversely, when we hit the mark, most of the time, that's not disputed. There's no moving of the goalposts.
With LiveLift, you also set a goal up front, an outcome that you care about. It might not be $0.35 CPID. It might be, "I need to move $10.5 million of incremental sales this quarter. Tell me the rough range of cost of that, and then that becomes kind of the mandate. That's the whole point is recording that system and knowing what we're solving for.
AI is clearly an accelerant for LiveLift, right? Talk to me two, three years, kind of what's your vision of what you wanna be able to build with something like this product where you think that that may be the end state?
Yeah. I think the vision is you are a brand manager and/or you are in the commercial team of a CPG organization, and you wanna know that you're using the best possible approach to Smart Promotions to inflect sales, and you understand the profitability characteristics and the trade-offs between profitability and volume well enough to go in on your own in a self-service tool and log in and, you know, configure a campaign based on where you wanna strike that balance. This bundle here is $0.30 CPID with $10.5 million of incremental sales. This bundle here is $0.45 CPID with $22 million of incremental sales. The system will configure in the background a recommended campaign for you.
Over time, as they see that working and trust that our system is optimizing the parameters of the offer, we're able to abstract away the number of choices that they have to make in that flow to where they can really just define the outcome and then let us deliver that outcome, right? Know that as they perform more and more and more campaigns, the thing is getting smarter. The model's getting tuned better so that we've run more experiments, we're more confident in our projections, and better able to, you know, to hit the mark, right? We don't hit the mark in every LiveLift campaign. We hit it in the great majority. You know, 100% is like a Soviet election, you know? Not everybody believes in it. We're hitting it a really high percentage of the time.
We wanna see that continue and get better, right? That's, that's our vision, and I think that, when you think about all the different buckets of money that are being spent without that level of sort of ability to monitor on a rolling basis and course correct, there's so many dollars that I think would rather go to work in a place with this level of accountability, where if it's not performing, you just turn it off with no penalty, right? You know, it's just like if you have a great Google campaign that's paying back at the right CAC, you leave it rolling and it's always on.
I mean, the vision is that people are running these things pursuant to the rules and constraints that they've set up, and we have greater visibility and predictability in our business, and ultimately, these rules, are ones that everybody's happy with. There's no debate about whether it's, you know, a critical part of their strategic approach.
If I think about the AI transition of commerce more broadly, right? The move to agentic. Understood so much is yet to be kind of determined, but what's your kind of vision for how Ibotta would fit in that future?
Yeah, I think there's gonna be multiple different places where people go to inform what they put in the basket. It could be that happens on a retailer's site through an agentic solution. It could be that there's some version of that in an LLM that stands in front of that. Even in that case, there will be some kind of fulfillment node, right? In both cases, as long as that AI is aware of the presence of discounts, it's gonna factor that into its decision of what to put into the basket, right? Price is a core element, so is a discount, is a core element of what a consumer cares about, which is value, right?
As long as value is a major kind of utility function for consumers, we're gonna be highly relevant whether they choose those offers or those offers are chosen on their behalf by agents. There are other forms of advertising that may or may not persuade an agent, but what we know is that price and value will always matter. I think that as commerce evolves, as more of it moves to online commerce, buy ahead, pick up in store, we know we're intercepting consumers all along that path of purchase, very high redemption rates. That's a tailwind for us, and I think agentic is just another piece of surface area where you will be able to encounter that content.
You talked about pricing a little bit on the last call, and, I guess the question is, talk about the pricing strategy and kind of the evolution of how you think about, you know, Ibotta's fee as it relates to the bigger picture of a campaign.
Broadly speaking, you know, we've always been pioneers in a positive way on pricing. We were the first to have fee-per-redemption-based pricing instead of fee-per-clip. That's a really big deal because brands don't wanna bear the risk of variable redemption rates. They want certainty, and they want to align our interests, you know. They want sales, not clips at the end of the day. There were a couple of instances, though, where our tiered rate card structure, didn't make as much sense for a client. For example, we would have, you know, if your product costs $0-$2, you pay this fee per redemption. If it costs $2.01-$3, you pay this fee and so on and so forth.
We needed to come up with something that didn't have these breaks, where if you were on the front end of that pricing tier, our fee as a percentage of the cost of your product was meaningfully different than it was on the other side of that tier. We wanted to create a uniform approach because that is more conducive to having a continuous function, is more conducive to a machine learning optimized solution in the future. There were situations where it was gonna be hard for certain clients to have profitable revenue unless we, you know, level set our fees. We're solving to maximize revenue, not maximize fee per redemption, right? Obviously.
There were some instances where we could take a step back and say, "Look, for instance, if you were willing to be always on with all your marquee brands, we would be willing to work with you on the fee so that you could get to the range of CPID that you really want, but in return we're gonna more than make it up in volume." We're doing those calculations, and that's the kind of thing we took a look at.
Okay. Talk to me about growing what is the publisher network, right? You guys have added some major grocers that are out there and delivery services. How do you think about growth from here now that it feels like demand is more stable?
I'm in favor of it. I think that we are a little overdue on, you know, announcing new publishers, but it's not because we're not making great efforts and strides to do that. Ultimately, we are always trying to add publishers because we believe that over the long run, the more publishers you have, the better your model gets, the more your data gets richer, gets more better predictive power, LiveLift gets stronger. The presence of that scale, the ability to move markets at an even higher rate, just means that we're impossible to ignore. You know, any CPG organization kind of has to have a strategy for how they think about intelligent promotion in the future.
Even if it might put short-term pressure on, the available offer supply, it's worth it because we see that step-up occur over time, even if there are temporary mismatches. You know, we think that the success we've had with Walmart Cash and just how much money people are making there, I mean, when you show a screenshot of a product and you say, "Here's, you know, Chips Ahoy! on Walmart, and you're getting $1 off, and here it is on your website, and you're not. And oh, by the way, your baseline price is probably higher than Walmart too.
You are way off. In this economy, Americans care a ton about the cost of Chips Ahoy!. You'd be surprised, right? They're not going to keep shopping in places that are just structurally more expensive, and there's no reason for you to do that, right? Why would you not pass along all this money and let it accrue to the benefit of your competitors when this doesn't cost you anything to join our network? We're finding that to be very persuasive. Not surprisingly, the predecessor companies that didn't focus on profitability and don't have third-party measurement and don't charge on a fee-per-redemption basis, have really struggled to get much content from these national budgets. The disparity between what we can bring and what they offer is manyfold.
It's not like, "Oh, we have 30% more offers." It's like, we have 4x their number of offers, or some cases even more than that, or way more than that. It just depends on each individual circumstance. We have a strong value proposition. It's just a matter of the decision cycle, the product roadmap cycle, the work to get it done to send us the data we need to power this transformational LiveLift product. I think they're starting to understand that that's gonna attract a lot more offer content that they want, so they're gonna wanna participate in empowering that. That, I think, is an exciting form of kind of synergy between the two sides of our business.
All right. Last question. You guys had 30% EBITDA margins in 2024. Right? You guys have gone through this phase and this transition. Talk to me about the path of getting back to those levels and kind of your desire to push on either additional investments to focus on growth or margin expansion.
Yeah. Look, I think there are a number of different ways of looking at that. Let's talk about just revenue growth. I mean, we envision being able to flatten out our cost structure over time. We've been investing heavily in R&D and innovation and, right, I think that's been really well worth it. But as we start to see more and more people use our Smart Promotions and invest in products like LiveLift, we should see the revenue increase and the costs not increase at the same rate. You know, we're contributing a high, you know, portion to the bottom line of every incremental dollar of revenue as it is. We don't have any capital expenditure. It's low single digits capital expenditure. So we should see that natural leverage in our business.
Then over time, you know, we're always looking to find ways of being more efficient, you know, in our business broadly. Like a lot of other businesses, you know, we wanna operate as cost effectively as we can. That's a part of how we will continue to manage those costs. Look, I mean, I think that the big thing here is we've had to invest out in front of the capability to unlock dramatically more offer supply, and you'll see the rebounding of those. Of course, the publisher negotiations, you know, that matters too in terms of our Adjusted EBITDA margin. What is the cost of sales of acquiring that deal? As we gain more scale and more leverage, we expect to have favorable deals, right? W e by and large have been able to stick to that up to this point, and I think that will continue.
Okay. With that, we'll leave it there.
Thank you so much.
Thank you.
Appreciate it. Good stuff, man.
Thank you.
That was great.