Fidji Simo: Absolutely. So we still have very high conviction on our long-term targets, which are that advertising would reach 4% to 5% of GTV over time. It will not be linear. If you look at the last couple of years, advertising has grown very fast over the last couple of years. And so that’s something that has changed in the last few quarters, mostly as we’ve told you before, because ad growth lagged GTV growth, and we haven’t seen ads fully catch up yet. But our conviction around the long-term model is very strong. The reason we have this conviction is because we continue adding more and more partnership to measure the performance for us and optimize that and the results always come back showing that we are an incredible place for advertisers to place their dollars.
You have seen that probably in the letter with our partnership Certana. We continue to launch new objectives and optimization capabilities like, for example, new to brand objective and target ROAS this quarter. And so all of these capabilities are really contributing to giving brands confidence that we have a channel by which they are going to see high return, high sales lead, high incrementality. In terms of what to pay attention to, in addition to these measurement improvements, targeting, etcetera which is really our bread and butter. We have extended the strength of our platform beyond the world of Instacart app. You have seen us do partnership with Google, recently NBCU, with Tradedesk where we are taking the data and the strength of our targeting of our audiences and applying it to advertising platforms outside of Instacart.
We have taken a similar approach with retailer ads where we power ads on our retailers’ website, whether that’s perhaps Schnucks many more. And that’s also a way to continue growing the ad business. And then finally, expanding HSA’s business inside the store on Caper with Caper Carts, which is obviously small right now given that we have hundreds of thousands up there, but as we ramp up to 1,000, of course, this is something that will get a lot bigger, especially as you look to 2025. So overall, like we feel very confident about the performance of Orad, some initiatives take time, but we feel good about our trajectory.
Colin Sebastian: Okay, thanks, Fidji.
Operator: Thank you. [Operator Instructions] Our next question comes from Doug Anmuth with JPMorgan. You may proceed.
Doug Anmuth: Thanks for taking the questions. Emily, can you provide any more color on what you’re seeing across cohorts? And if you see a timeframe for mature cohorts, those declines to flatten out? And then anything that you can help us understand around the benefits of the new partnerships with Kroger and Costco on EBT SNAP and on quantifying just as you flip from a headwind to tailwind over into 2Q? Thanks.
Emily Reuter: Thanks for the question, Doug. So on cohorts, what I’d say is that everything that we’ve said about cohort dynamics remains true. So a mature cohort declines continued to improve in the quarter, and new cohorts are bigger than pre-pandemic. What I would note, though, is that the strong Q1 performance that we saw benefited all of our cohorts. And so for that reason, I would expect that there was some acceleration and improvement in Q1 that we don’t expect to repeat in Q2 given those one-time impacts. So overall, I think the underlying trends continue to move along the same trends with the one-time impact in Q1. As it relates to the benefit of the new partnerships on EBT, overall, what I’d say is that the year-over-year headwind improved from Q4 to Q1.
So the impact was strongest in Q4. And in Q1, the launches of Kroger and Costco helped mitigate that impact. As we move into Q2, we do expect to lap the expiration of those benefits. But the overall impact is modest. The way that I think about it is that the EBT SNAP benefit in Q2 would effectively offset the impact that we had in Q1 from leap day, so about 1 percentage point of growth.
Doug Anmuth: Thank you.
Operator: Thank you. [Operator Instructions] Our next question comes from Michael Morton with MoffettNathanson. You may proceed.
Michael Morton: Good evening. And thank you for the questions. Two, if I could, they’re related. Could you speak to the defensibility and contract length of some of the largest and exclusive relationships. It’s a question that we frequently get from the investor base and any clarity there. And we do appreciate that it’s not essential to the success of your business, but it’s FAQ. And related to partnerships, I would love to know how you think longer-term when it comes to working with the likes of Walmart and Amazon. Research shows that they’re taking share in grocery and could be taking share from your core customer base on the grocer side. And a lot of times, they like to pull things in-house eventually. So is there any risk that you’re feeding the beast and bridging them to their own sustainable solutions. Thank you so much.
Fidji Simo: Thanks for the question. So on contract length and exclusivity. First off, I just want to remind everyone that exclusivity is not our strategy. The strategy is fundamentally to be the partner of choice by being very embedded in grocery business because we drive growth for them. As for the contracts, they are, I would say, 1 to 2 years, and they all have different timing. So it’s not as if this cliff coming up at any point. And what we usually see is that when retailers go non-exclusive and go to other platforms, what happens is that they continue to grow with us, invest more in the relationship and the use case that happens on other platforms is different from our own. It ends up being very small baskets, the thing that you add after the restaurant order like a bag of chips and can of Pepsi, so very different from what we can do on our platform, which is really everything from small baskets to large basket.
And so we continue to invest in our technology with grocers. We have a relationship even when they’re not exclusive, where very often, we power on an operated business, we power our fulfillment. We do kickup with them. We do EBT SNAP. We do virtual convenience, the lease goes on, and it’s obviously a much more strategic relationship than just sitting on the marketplace. So that’s very much how we think about this. In terms of partnership and long-term working with competitors, I would say what we obsess over is always having the best possible selection for customers, and that’s what got us to have a selection that represent 85% of U.S. gross retail being on Instacart. And so that’s why we want to work with every grocer. However, when you talk about competitive advantages, we think our competitive advantages are very strong and in particular, when it comes to fulfillment advantages, the scale at which we operate gives us very significant advantages over any retailer wanting to do that themselves.
And that’s why you are seeing even the largest guys, the Krogers of the world, etcetera, really relying on us for fulfillment because we can do that at economics that are highly competitive at the scale that is incredibly competitive with quality that’s top notch. And as a result, we can grow their business rather than doing that themselves and not growing as fast as they would if they relied on our technology. So we want to make it always the smartest choice for them to partner with us, and that’s exactly what has happened to date.
Michael Morton: Thank you.
Operator: Thank you. [Operator Instructions] Our next question comes from Andrew Boone with JMP Securities. You may proceed.
Andrew Boone: Thanks very much for taking my question. The NBC Universal partnership is very interesting. Can you talk about the potential of off-platform advertising and what you need to do unlock more budget as well as more partners? Thanks so much.
Fidji Simo: Thanks for the question, Andrew. So the way we think about it is that we have really unique customer data that can help advertisers identify audiences, for example, we would be new to their brand, who is an engaged customer that may – that used to be engaged, but may have churned. All of these audiences are incredibly critical for advertisers especially in a world where cookies are going away. And our other platforms are really hungry for high-quality third-party data. And so we are able, through our first-party data to do these partnerships where advertisers are able to take our audiences and leverage that on other platforms to make their advertising on other platforms more performant. And so that’s very exciting.
In addition to that, some of these brands decide to point directly to Instacart because they know that people are going to convert that if they land on a page where the ads allow them to buy and get the product inside their hands within an hour. So it’s really a full funnel approach that we have with Google, the Trade Desk and the [indiscernible]. And that’s why we’re really excited to continue deepening these partnerships. Now in terms of materiality, I would say it is still small because it’s a new thing and it requires a new infrastructure and a new muscle. Very often, the people that we talk to, to go get these budgets from different people than the people that are allocating spend on Instacart because they’re just allocating spend on other platforms that are more maybe awareness-driven or other objectives.
So it takes a bit of time to settle this in, build these new relationships, add the new tools to make these services easier to access optimization capabilities, but we’ve done that in the past with our own platform. We know exactly the playbook, that’s the playbook we’re applying to this year, and we hope to see the results of that in 2025.
Andrew Boone: Thank you.
Operator: Thank you. [Operator Instructions] Our next question comes from Ron Josey with Citi. You may proceed.
Ron Josey: Hi, thanks for taking the questions. Maybe as a quick follow-up to one of your prior questions, Fidji. I wanted to ask about just the comments in the letter around operating data. I think you had and you talked about a decade of operating data. And I wanted to just understand the strategic moats that you have as a result of that, both for the consumers and your retail partners, because I think that’s something that’s often discussed. And to that, maybe another question around, I think the comments on the call around quality – or in the letter around quality, specifically around receipt analysis and the benefit of LLMs. Just talk to us about the benefits that you gained from that. Thank you.
Fidji Simo: Absolutely. Thanks for the question. The reason I talked about operating data in the letter is because very often, we talk about having the best prediction models and things like that, which is incredibly important, I want to be clear. But even if someone literally duplicated, right now, all of our prediction models if they didn’t have the 10 years of operating data, the model wouldn’t be that accurate, only have small part of data you collect. And so what we have done is really amass a large pool of data, both on the consumer side and on the inventory side. On the consumer side, it’s 11 years to 12 years of large basket data of everything that customers want to buy, everything that they substitute, which is critical because that allowed us, if an item is not on the shelf, to substitute it with something that they prefer.
Lots of data about the types of products that are complementary to others that we can surface the best product recommendation and also surface the best ad recommendation. So, all of that contributes to a richness of consumer data. That means that when you have invested in Instacart, you have placed a couple of orders, it’s really hard to go to another platform because you would have to rebuild your entire basket or rebuild all of your habits, rebuild all of your preferences. And we have all of that for you. And that’s why you are seeing in the letter, we mentioned that 75% of our customers have purchased an item from the Buy It Again list because again this is data we have on all of their purchase history. Now, on to kind of the shopper side and inventory side, it is also really critical to have data that allows us to predict what’s on the shelf.
And so again, because we have 600,000 shoppers roaming the out of grocery stores and we partner with retailers to get not only their catalog files, but their balance on-hand data in some cases, we are able to create a system by which we can predict quarter-after-quarter, what’s going to be on the shelves with more accuracy, given that we have developed all of these technologies to tell us what’s exactly on the shelf right now. And as I have mentioned in previous calls, the biggest complement that we have on our technology is that some retailers and CPGs are actually using our out-of-stock predictions to optimize their operations because we know what’s on the shelf much more in real time with much more accuracy than retailers and CPGs themselves.
