All right. While I pound my coffee over here, we're going to get started. Thank you, everybody, for attending. Ross Sandler, I head up the U.S. Internet team here at Barclays. We're very excited to have the debut of Instacart as a public company here. First time they've attended a conference. So Nick, thank you.
It's a pleasure to be here.
To start, before we start, I have to make a disclaimer. Want to note that Nick might make forward-looking statements about Instacart's performance and expectations. Those statements are subject to risks. A full list is available on Instacart's latest 10-Q. Nick may also reference certain non-GAAP financial metric, metrics and reconciliations, which are all available on Instacart's IR website. Okay, so with that out of the way, Nick, grocery is just this massive TAM, but it's also one of the lowest online penetration markets. Why is that? I mean, we're ten years into it. Why is it still such low penetration?
That's a great question, and you're right. Grocery is a $1 trillion-plus market in North America, and the online penetration is 12%, even post-COVID. And to put that in perspective, the $1 trillion market is about twice the size of the restaurant market, and the online penetration at 12% is about half the penetration of restaurants in North America. Now, the good news is that people expect, and we believe, that online penetration will double or triple over time. But the question is, why is it moving slowly? It's moving slowly because it's very difficult to do. It is very different than restaurant food delivery. It starts with the level of the enterprise integration that you need to have with the grocer. Grocers have hundreds of locations. Inside each location are tens of thousands of SKUs.
Those SKUs have different prices by location, and the prices change over time. The availability and the inventory changes frequently, and so you need to make sure that your catalog is correct. You then also need to find ways to integrate with loyalty programs to make sure that you're able to pass through savings to customers, to integrate with EBT SNAP and other forms of payment, to make sure that you can replicate the offline in-store experience online, and that's just for the retailer part of it. Once you do all of that, you need to make sure that the customer can get what they want. The way that people buy groceries in North America is, it's anchored by the weekly shop. You know, they buy big baskets once a week.
So to create a big basket, that could be 20-25 items, you need to make your app or your software functional to help the customer pick the right 20 items out of tens of thousands of SKUs. And that can be really cumbersome if you don't have the right UI, the right buy it again recommendations, the right potential replacements for things that are out of stock. And we take our data, and we overlay it on top of the catalog data to make sure that we can highlight for consumers what's likely to be plentiful, what's likely to be out of stock, what prices are really attractive, where are there savings. And we can do all of that while also providing an opportunity for brands to compete to get in the basket.
All of the advertising that happens on Instacart happens when a consumer is building their cart, and they're searching for cereal, they're searching for a beverage, and brands can compete to be the one that's added to their cart. That's really important because it allows us to have much lower fees than restaurant food delivery, about half of the merchant and consumer fees, because we have this diversified model and larger baskets. Then the work that the shopper does is very different. A shopper is not just driving from point A, picking up a bag, and taking it to point B. Half the time spent is spent in the store shopping, shopper communicating with the customer, making sure that they can find a replacement if something is out of stock.
And we do that more than once per order on average, and with 95% satisfaction of the replacement. If you do not have a good replacement, then the recipe might be broken, and if the recipe is broken, then the whole cart might be broken. And if you have that, then you might need to appease an order. So if anything goes wrong at all, you're stuck with a business that doesn't have good recurring consumer behavior and has high appeasements and refunds, and if you get it right, you can tap into really high-value customer activity of spending $100 per week on a recurring basis, and you can build a really nice business.
You mentioned a few of the competitors. We actually had DoorDash up here yesterday, and they actually espoused what you just said, how difficult it is to get the large basket right and how the error rates are much higher. But I guess the question we get from the investment community often since the IPO is just the competitive landscape. So how do you guys view the competitive landscape? And, you know, is it about competing? Is it about just driving more adoption? How do you think about that?
Yeah. So we're the clear leader in grocery delivery in North America. We look at third-party data, and we've talked a little bit about big baskets and small baskets. In big baskets, defined as $75 and above, we have greater than 70% share. In small baskets, $75 and below, we have greater than 50% share. And if you were to plot that share over time from before COVID to now, our shares increased. So we're doing really well in small baskets, and we're doing really well in big baskets, and it's because we focus on a couple of core things. One, making sure we have the best selection, and our selection is more than 1,400 retailers.
If you look at those retailers, and you look at all of their offline sales, they represent more than 85% of grocery volume in North America. So we cover 97% of households in North America. And not just having the grocer online, but having multiple services per grocer, having a pickup service, having a virtual convenience service, being able to accept EBT SNAP, being able to have loyalty integration, prescription solutions. Then we have the ability to really drive affordability. 425+ of those retailers are at price parity, meaning they charge the same price in store as they do on Instacart, and retailers set the price on Instacart. Instacart does not set the price. Beyond price parity retailers, we have discount retailers, we have club retailers.
We want to make it easy for a consumer to find the store where they can, where they can save. And then convenience. Convenience, for us is all about making sure that it's a huge part of our value prop. We help people save money. In the most recent quarter, 36% of our orders were priority orders, where you could get a full big basket grocery experience delivered in less than 50 minutes, and that just costs an extra couple bucks. So for all of those reasons, in addition to quality, we talked about making sure that you could get an order accurate. We look at found rates, we look at fill rates, and our levels for found rates and fill rates are at higher levels than they've been since 2019.
At this mega scale, they continue to get better.
So I guess then the question becomes, if penetration's low, competition is, you know, in the distant rearview mirror, if you will, why aren't you guys growing faster? I think that's a question that many investors are kind of curious.
Yep, it's a great question. We're really focused on continuing to accelerate order growth. The good news is that we have accelerated order growth every quarter this year from Q1 to Q2, again from Q2 to Q3. We are doing really well competitively. As we just discussed, we're not losing share. We continue to see ourselves gaining share over the long-term horizon. And what we see is that we're really good at acquiring new customers. We disclosed in our IPO, if you just look at the brand-new customers that started shopping with Instacart in the first half of 2023, that was more than 60% larger than first-time customers in 2019.
That's the relevant comparison period, because there was a lot of COVID activity in 2020, a medium amount in 2021, and still a small amount in 2022, although it was really concentrated with that Omicron wave in the first quarter. And that suggests what's happening to us right now, which is there are a couple of headwinds impacting the business. Specifically, you know, the growth equation is mature cohorts ideally grow, and then you add new cohorts on top of that, and you add those two things together, and you have really nice growth. And our mature cohorts grew in every period, year-over-year, with the exception of the big 2020 COVID cohort. But what we're seeing in 2023 is that the mature cohort shrank in Q1.
It was the first comparison of a sort of no COVID period back to the Omicron wave that was happening a year before. But those mature cohorts improved. They still shrank, but in that, improved in Q2 and then still shrank, but improved in Q3. If you were just to take the two large cohorts that we have, 2020 and 2021, they were slightly over 50% of our business last year, slightly under 50% of our business this year. But if they're declining, it's a huge headwind to growth. If, if you eliminate those two cohorts, which we don't do because they're a critical part of our unit economics and scale, the business is growing more than 20%. If you were to say, like, "Can you get those two cohorts back to flat?" Then the business growth rate doubles.
So we really need to focus on making sure that we re-engage, resurrect users from those mature cohorts in order to structurally change the math of our growth rate. We're pleased with what we're seeing from new customer acquisition. We just need to impact. We just need that headwind to stop being a headwind. And then there's another headwind that is not getting better throughout the year. It's actually gotten slightly worse throughout the year, and that's EBT SNAP volume. EBT SNAP is government food assistance. We pioneered along with our retail partners, bringing those benefits online, and that's really important. But when we did that, we built a really big business of EBT SNAP, and then in Q1 of this year, emergency benefits were cut 30% and post-COVID, and that was really tough for our consumers.
They had less money to buy groceries, and that impacted us. So what we're seeing this year is, in 2022, we continually grew EBT SNAP volumes throughout the year, and then in 2023, it dropped. And it's been relatively stable, but the comps keep getting harder and harder, and so the year-over-year impact is getting worse throughout the year, and we should lap that at some point early next year, but it has been an increasing headwind for us this year. So structurally, we feel good about competition. We feel good about acquiring new customers. We have identified the headwinds that are facing our business, and we just need, you know, time plus execution to get past it.
And you mentioned the importance and the difficulty of large baskets and the large basket order. I guess, what competitive advantage does that provide you? And, you know, you guys also talked about how it took, you know, 100 million orders before you actually turned gross profit positive on a per-order basis. So how did large basket help with that?
Yeah. The most important reason that large baskets are important is because that's how consumers shop for groceries in North America. So you have to build a product that serves that use case. Small baskets are also important. It's easier to activate with a small basket. It helps drive additional engagement. So small baskets are also important, which is why we're really pleased that we have greater than 50% share of the small baskets. But big baskets are critical, mostly because of that consumer use case. More than 75% of grocery sales in North America happen with big baskets, so you've got to be there if you want to have success. Now, it's much harder to get accuracy across 25 items compared to two or three items. It's much harder to build the basket.
It's much harder to fulfill that order, and so you've got to attack every part of that, of that chain with technology and with experience, and that's what we've done. It also allows you to have lower fees, because it costs money to send a shopper to a store, and it costs them money to send a shopper through the aisles, and it costs money to send a shopper to wait in line and check out, and then it costs money to have them deliver the order, and whether it's a $30 order or a $100 order. So the key is finding ways to do that for the big orders... Then once you have the density to be able to have the shopper shop for two orders at once, that's what we call our batch rate.
Batching is key to the unit economics and the improvements that we've seen in our business over time. Compared to pre-COVID to now, we've had a 50%+ increase in batch rates, so number of orders per batch. So the key driver of efficiency and cost to serve is not the number of shoppers that we have or the number of consumers that we have. It's the number of big basket orders that we have happening in the same store at the same time, because then you can send a shopper to the store once, they can navigate the aisles once, they can check out once, and then they can deliver as two nodes on a delivery route. That means that the shopper can have a better earnings opportunity.
They can earn two tips, they can earn a larger batch fee, but it reduces the cost to serve for that order.
And you mentioned that even in small basket, you've got really good share. I guess, how do you view, you know, small basket, and how do you view that in the context of expanding your customer base and your TAM?
It's a great point, and it's an important question. We think it's easier to go from big baskets to small baskets. We think that small baskets are really helpful in acquiring new customers because it's easier just to buy a couple of things than convert. It's then very difficult to turn a small basket customer into a big basket customer, 'cause you have to go through all of the same challenges of building a very big basket and fulfilling it. So we look at third-party data, and it says, you know, compared to some of the other new entrants in the category, we activate multiples of the total GTV that they activate. For big baskets, we're five times better at activating big baskets, more than five times. And for small basket to big basket conversion, we're also more than five times better.
Those customers that start with the small basket, what percentage of them ever make it to be big basket customers? We have real advantages in the ability to do that. Small baskets are really important. They're an important part of growing the user base, engaging the user base, retaining the user base, which is why we're really pleased with our share in small baskets, too.
Okay. And then if we look at the customer base itself, you've got just under eight million paying customers. The typical, you know, large scale restaurant delivery app might have maybe 20-ish in the U.S. So you guys trail that by a decent clip. I guess, what's more important, growing that eight million or increasing frequency within that eight million?
Yeah, it's a great question. So the way that the math works for a customer is different for grocery delivery than for restaurant delivery. Fewer customers, but customers spend much more, but they start spending much more sooner, and they expand less over time. I can just give you an example. Based on the cohort disclosure that we provided, a customer for us, an active customer in their first year would spend more than $200, and by year six, they're spending more than $450 per month. And, that's great. There isn't another, you know, service that we've seen that has that amount of spend per customer, but the expansion is less.
You could compare it to starting by buying a $125 order, and then, you know, you can expand by eight times. You're still just approaching where we're beginning in terms of the total spend per customer. So we have fewer customers, but more valuable customers. And the more valuable customers expand a little bit less over time, but they start already at a level that is very, very high. So for us, the critical thing is continuing to add more customers, and that's what we're focused on doing, making sure that we acquire new customers, which we're doing a good job of, retaining them, resurrecting customers that have tried us in the past but have gone back to the store post-COVID.
We have a very long-term view, and that's our primary focus is increasing the number of users and orders that we have.
Okay. Shifting gears to food inflation and AOV. So the last two years have been a bit of a roller coaster. You've had food inflation, now we're in kind of a food deflation mode. How does that impact unit economics and frequency, and how will AOV trend go forward? Your guidance is more predicated on, you know... Is it more predicated on GTV or order growth?
Yeah, great question. So just to set some context, before COVID, our basket size was around $105. During COVID, in the peak of COVID, it was up to more than $130 per order. And then it steadily declined as people were no longer sheltering in place and buying everything that they needed, online, and it got back down to around $106, $107. And then we started to see food inflation, and it's climbed up, to around $112, $113. And so, that has sort of a direct positive short-term impact on our business as basket sizes go up, because we earn fees based on the baskets that we sell, and it's easier to fulfill orders that have fewer items in them.
But it also has this indirect pressure, right? Like, the number one thing that consumers in North America think when they think about groceries over the last couple of years is they're too expensive, and so they're looking for ways to save. So we actually believe that inflation decreasing or going away might have an impact where basket sizes shrink or more stay flat instead of growing. And if that happens, that would not benefit our overall GTV growth directly, but indirectly, it would make consumers' perceptions of affordability improve. And it could be a positive thing for online grocery adoption over time.
So what we've been focusing on over the course of our last couple of calls is the difference between order growth and basket size growth, and we actually think higher quality growth is order growth because of this potential for basket size trends to change over time as inflation changes. And order growth has been accelerating at a faster rate than GTV growth throughout the year.
And you talked earlier about, you know, the how complicated it is for shoppers to go in and get these big baskets, you know, set up in the right way, and how driving higher batch rate and efficiency is important. If we look at the performance on your transaction take rate and your fulfillment efficiency of those shoppers. In the recent quarter, it was up like 80 basis points year-on-year from fulfillment efficiencies. I think you invested about half that back in the way of consumer incentives. So I guess, could you just talk about where we are with efficiency, and then how do we balance reinvesting some of that upside back in, and why, you know, the 7% range is the right range for transaction margin?
Yeah. So our transaction revenue is three components that are positive revenue and three components that are contra revenue. The positive revenue components are retailer fees, consumer fees, and payment revenue. The contra revenue components are what we pay shoppers, shopper pay, consumer incentives, and then appeasements and refunds. And the biggest driver of improvements in transaction revenue over time has been efficiency gains. That has reduced cost per order, even as earnings per shopper have increased over that same time period. And so the question is really about balance. What is the right level, and how do we make sure that we can reinvest those savings into incentives or fee structure, or creating new services? And so that's why we're at 7.2% transaction revenue.
As a % of GTV, our long-term range is 6.5%-7.5%. People ask the question, like: Is there more room to go? Can you get higher than 7.5%? We probably could, but that's not our strategy. A good example of that is we like to reinvest, and consumer incentives continue to grow the business, and we like to reinvest in product experiences. We spent a little bit of time earlier talking about priority delivery, which is paying us a couple extra bucks to get your order sent right to you less than 50 minutes. We also innovated last year with something called No Rush Delivery, about a year ago, which if you give us more time, like a two- or three-hour window, we can save you money, reduce your fees.
And the way that we do that is if you give us a two- or three-hour window, there's a greater opportunity for us to batch that order, shaping demand into those windows, batching multiple orders, which reduces the cost to serve. But we don't flow that through to profit, we give it back in the form of discounted delivery fee. And we'll look to continue to do things like that, where we are finding ways to reduce the cost of our service, but not just by reducing our fees, but making it more efficient and then passing those savings through to consumers. So it's a balance, and for that reason, we think that 6.5%-7.5% is the right range.
Even though it could theoretically be higher, we don't think that taking it all to the bottom line would be the right balance.
Okay. And similarly, on the advertising side, on the ad take rate, I guess stepping back, the advertising business has been a major differentiator for you guys as far as the overall story. As we look forward, we're kind of halfway to that, you know, 4%-5% of GTV range. What are the two or three big things that you have in advertising that you're working on that might bridge us from here to, to that goal?
Yeah. So we look at advertising and other revenue together as a percentage of GTV, and we call that our investment rate. And we think that the investment rate obviously started at zero and has gotten up to, you know, high twos, and the long-term target for that is 4%-5%. It's first important to understand why the target is 4%-5% and not 5%-8%. You know, we get asked the question, there's other marketplaces that are already at 5%, on their way to 8%. Why is that different for you? The main reason is that not all of our GTV is as advertisable as the rest.
Now, if you think about a grocery store, there are certain aisles that have higher investment rates, like packaged food or beverage or alcohol or personal care or home care or pet. And then there are certain aisles that have lower opportunities and lower investment rates, like fresh produce or fresh meat. And a good way just to think about this is if you do a search for light beer, you're likely to see more ads than if you do a search for lettuce. It just, it makes sense. For those reasons, on a blended basis, we think the right long-term target is 4%-5%. So how are we increasing how have we and how are we going to continue to increase the ad investment rate? It's by demonstrating, incremental growth and, and ROI for our advertisers.
And we look at our advertiser base across various tiers of this very large strategic accounts, the growth accounts, and the new and emerging accounts, and we make sure that we continue to innovate and show them new ad products, but also new measurement and campaign management capabilities, and A/B test to prove the incremental sales lift that we can provide. And we've published a number of case studies that show consistent incremental sales lift of, you know, 15% plus, and in some cases, much higher than that. And so by continuing to show really high returns and continuing to show that we're a channel that is growing faster than the overall grocery industry, we tend to attract more budgets because it's highly measurable bottom of the funnel.
We'll need to support that with new innovations on products and that are consumer-facing ad products, and also new innovations on measurement and campaign management. Our long-term target is that we want to see about 20 basis points of improvement on an annual basis, but it is lumpy and it is, you know, for the last five quarters or so, it's been higher than that. It's been 30-50 basis points, and for the next couple of quarters, we guided that it would be lower than 20 basis points. And that's because we were comping, starting in Q4, and for the next few quarters, against a very, very successful expansion, where greater than 35%—35 basis point investment rate increases because we launched at this time last year, shoppable video, shoppable display, and new campaign management tools.
So that for those reasons, we do see consistent ability to improve ad investment rate over time, but it will not be linear.
Okay. And as we go down the P&L, your, your EBITDA margin pretty much inflected with the rest of the business in 2020. You guys invested into that. We're about halfway towards the 4%-5% goal as we sit here today. How are you guys balancing out, you know, growing top line versus seeing leverage on EBITDA?
Yeah. So our top priority is profitable growth. We want to be able to grow a healthy, sustainable business that can produce returns over time by balancing all four sides of the marketplace. And if you look at the first point just to make is, all, all different revenue types are all high-quality revenue. Our corporate-level gross profit margins are in the mid-70s. So we are, we are not, it's not like an additional $0.25 of ads is higher flow through margin than additional $0.25 of savings or an additional $0.25 of consumer fees. All of that flows through at similar rates. So but how do we get from where we are, which is in the, you know, low twos, to our target of 4%-5% over time?
We see the additional ability to continue to expand our revenue margin, especially in ads going from, you know, high twos to four-five. And then we do expect to see operating leverage over time. We have invested significantly in the scale of our business and our R&D. We're a technology company. We have a large R&D team. We think that the current size of our team can continue to support our growth aspirations, and so we would expect to see, you know, improved operating leverage in addition to the revenue margin improving over time.
Last question is on capital allocation. So having done lots of IPOs in my era, you've done many in your era as well. Like, you guys actually announced a share repurchase program fairly close to the IPO. So thoughts behind that, and just overall, as you look out, you know, yeah, what's the overall thesis around capital allocation?
Yeah. So we have at the time of the IPO, we had more than $2 billion of stock on the balance sheet, and we did not want to do an IPO that would raise money and dilute the company because we already had enough money and we continued to produce cash. So we did an IPO that was anti-dilutive, which means the day after our IPO, we had fewer shares outstanding than the day before our IPO. And we did that by having secondary shareholders and former ex-executives and founders sell secondary stock in the IPO. And the primary shares were essentially the stock that had already been issued under our RSU programs, where we just sold those to pay the taxes that were due at the vesting.
Consistent with that belief that we want to do things that are anti-dilutive, and we want to take advantage of our large cash balance, we thought it was logical, on our first earnings call, to announce that we intend to be a share repurchaser over time, and we announced a $500 million share authorization. We also wanted to make it clear, like, we know our float is small. Our intention is not to make it harder for investors to buy our stock, but we are a buyer, an opportunistic buyer of our shares over time, and we wanted to make sure that we had the share repurchase plan in place ahead of the upcoming lock-up release, and that we were signaling from the beginning that we intend to be an opportunistic buyer of our stock over time.
So that's what was behind that announcement.
Great. So with that, we're going to end this. Nick, thank you very much.
Thank you.
Congrats on all the success. Thanks for coming and,
Appreciate it.