Elliot Turner of RGA Investment Advisors presented his in-depth investment thesis on Grubhub (US: GRUB) at Wide-Moat Investing Summit 2019.
Elliot calls one of his favorite setups “the post-hype sleeper” after the fantasy baseball term for a player who joined the league with considerable hype only to take longer to develop than the average fan expected. Markets are pretty similar in how early hype gets extrapolated too far, and when reality settles in, investors must grapple with whether the setback in valuation is temporary or permanent. Such stocks get stuck in-between a growth and value investor base and are thus perfect for scouting out GARP opportunities.
For the second time in less than four years, Grubhub has set up accordingly as a result of a business model evolution from a pure two-sided marketplace connecting diners with restaurants to a three-sided marketplace adding in the delivery component. For the first time in over two years the stock is trading at a market cap to Gross Food Sales below one. With the nationwide rollout of turnkey delivery, investors are questioning the long-term margin profile of the business; however, Elliot believes margins are the wrong focal point. Delivery is inherently lower margin as a result of higher take rates. EBITDA/Gross Food Sales is more appropriate and has been pretty consistent around the 4.5% level, though with the nationwide rollout of turnkey it will be lower in 2019. For delivery marketplaces, network effects matter most on the local and/or regional level and Grubhub dominates the most delivery-friendly markets in the US. Grubhub boasts a large and growing addressable market, with low penetration, a smart management team proven adept in execution and disciplined in M&A, and a reasonable valuation that supports a DCF valuation into the mid-$80s per share.
Replay this LIVE session:
The following transcript has been edited for space and clarity.
I love these conferences. Not only do I get great ideas from all of you but also outstanding feedback which helps sharpen my own ideas.
My new idea is Grubhub. Before I presented Twitter in the spring of 2017, I had been contemplating whether to go with Twitter or Grubhub. Grubhub had a good run heading into that period, so it wasn’t quite as cheap. We’re GARP investors, so I focus strongly on valuation as well, and I felt I missed it.
First, disclaimers: we do own shares in Grubhub. We may sell them at our choice, but we tend to be long-term oriented.
To give a sense of our long-term orientation, this is the eighth idea I present at an MOI conference since 2015. We are now crossing the threshold where more than half of our portfolio has been presented. We haven’t sold a single share of those eight ideas. In fact, we own more of most of them than at the time of the respective presentation.
Before I deep-dive into Grubhub, allow me to introduce something conceptual. I like thinking in terms of mental models and their portability from one domain or discipline to another. Charlie Munger tells us to apply the big ideas from mathematics, physics, chemistry, whatever it may be, but I want to tell you about a really powerful idea from fantasy baseball. It’s not quite physics or chemistry, but I think there are some interesting concepts we can learn from fantasy baseball. In many respects, baseball, especially fantasy baseball, is similar to investing. Think of your team as a portfolio. Your players are your equities. Everyone has the same resources to analyze a growing trove of data. There are, however, constant inefficiencies for managers to discover. My favorite setup in fantasy baseball happens to be the post-hype sleeper, and it’s well-illustrated by Javi Baez, who, like Grubhub, is based in Chicago.
Baez was called up by the Cubs in 2014 to much hype, as a consensus top 10 prospect. Just about everywhere you looked, he was THE guy to go after. However, Baez really struggled in his first stint in the big leagues. In fact, he was a below-average player. It took him two years to start getting regular at bat, and at that point, it was as a utility man. From there, it took another two years to become an MVP-caliber player. In the intervening period, all the hype was gone, and Baez could be drafted in late rounds or in auctions for dirt-cheap prices. It’s similar to what we’re looking for in the market.
Let’s translate the post-hype sleeper setup from fantasy baseball to the market. Often, it comes in the form of a busted IPO, but it doesn’t have to be a recent issue. There’s rich valuation metrics, positive fundamental momentum along the way, and growing intrinsic value through key growth KPI slowdown, which leads the stock price to a sharp breakdown. Downside momentum overshoots though the business keeps growing in intrinsic value all the while. The stock is not necessarily cheap enough for traditional value investors, and growth investors are too recently burnt to come back. That’s where GARP comes in. You’ll often hear something like, “I love this company, and I’d buy this stock higher if numbers re-accelerate or lower if things stay the same, but not right here.” It’s a quote I hear a lot from different kinds of investors. Then, the stock settles into a range of apathy.
When Grubhub IPOd in 2016, the stock was priced at $26 and opened at $40. In many recent hot new IPOs, people are saying this is indicative of the top, whatever it may be, but these first-day pops are a pretty common phenomenon. Grubhub traded up from there before collapsing and shedding 2/3 of its value. We might have our new crop of future post-hype sleepers.
The company spent almost half a year building a base and then blasted off after that first decline. We started our position in this busted phase. While it had a great run, after recovering, the path was hardly linear. There were two individual days where the stock dropped 15% on new breaking out of Amazon’s ambitions in the delivery space. In total, the stock had four individual days of 15% decliens along the way. The lesson here is that the path to greatness is not linear.
I first intended to present Grubhub in the spring of 2016 at the Wide Moat Investing Summit and again in the spring of 2017, when I went with Twitter. In 2016, I presented on Envestnet. Each time I was planning to talk about Grubhub, the stock ran up heading into the conference date, so I had to pick something else. That was a good GARP buy, but let today serve as proof that there’s always another chance, both in terms of finding something to present and as long-term investors. I might have been out of ideas already had I come to this conference having presented Grubhub in 2017 considering how much of our book has been presented that we still own. It’s interesting to think about it along those ways. This year at Best Ideas, I already had to go back to the well with PayPal. Speaking of PayPal, I love thinking about portfolio synergies. One of my favorite things to do is order a Grubhub meal while checking out with PayPal’s One Touch solution. Each of my companies gets to earn a little margin on my behavior.
I’m big on the idea that there’s always another chance, which brings me to the heart of my philosophy on position management. I want to highlight a couple of key concepts I look at myself. I like the idea of imposed patience. With a long timeframe, we have the luxury of never saying, “I missed it.” After all, every investment is a hypothesis we’re testing along the way based on a decision made in an uncertain world. We get new information every day, and we have to weigh that information against what we already know. As a result, I never buy my full position size on day one. Given the idea that every investment is a hypothesis, the most important thing to know is when you’re right and when you’re wrong. We want to find those opportunities where Soros says something like “Go for the jugular.” Personally, you can get far greater conviction in something I know well, something I’ve seen move in the market, than something newly entering the book. One big reason behind that is in a complex world, the longer you’re involved in something, the more you can simplify a thesis. If you get to that point where you can simplify a complex thesis with just one variable, you’ve got something quite special.
I want to start this story with the merger of Grubhub and Seamless. It combined two early competitors, creating a two-sided marketplace in the delivery space. With the merger, the competition between the two became an afterthought, and they were able to consolidate their presence, expand to new geographies, and focus intently on increasing the depth on both sides of the marketplace – diners and restaurants.
Very early on, Grubhub had a 12% take rate for matching diners with restaurants. Demand was highly incremental for restaurants and helped push the range of delivery offerings beyond genres like pizza and Chinese food, which are the core offerings in a given geography. In the early days, all the restaurants were moms and pops, no chains, and fax was the way they received their orders.
This was before the iPhone and the app itself took off to allow diners to start their search and discovery within Grubhub instead of Google or anywhere else. It happened because Grubhub removed two crucial pain points for diners. First, most people have a drawer in their kitchen for the menus from a few delivery options in the area. Grubhub gave you a centralized place to know exactly which restaurants deliver to your address and what was on the menu. Second, calling in an order was prone to error in high-volume restaurants. With a neatly printed order sheet, the risk of error went way down. In the early days, when Grub boasted a 12% take rate, some loud observers were calling the take rate way too high. However, core marketplace take rates have risen over time to above 15%. This is because the company has become the starting point for diner search and discovery.
There is a huge total addressable market TAM), which is one thing I consistently look for. To break down the addressable market, we started with the total spend of food away from home. It is about $250 billion, with $70 billion of that pie going towards takeaway and delivery, and $30 billion of the takeaway and delivery slice already going to digital platforms, including Grubhub, DoorDash, Uber Eats, and Postmates and pizza companies like Domino’s, which are already pretty advanced in digital ordering. For the longest time, food at home and food away from home ran in locked step, but about 10 years ago, they started diverging, with food away from home taking the lead. In the last five years, its lead has meaningfully accelerated.
What you have, effectively, is a growing TAM. Delivery is becoming increasingly popular, and much of that is due to the success of Grubhub in removing the frictions of ordering. Demographics play a big role in this, too. The habits of young and old are diverging meaningfully. Millennials have strikingly different experiences and expectations for where and how to eat. In fact, the youngest cohort of Millennials have grown up with online food ordering to the point where it’s second nature – food away from home is much more important for them than for any other cohort.
One thing I’d point out is that success has its downsides. As well as things have gone overall for Grubhub, it hasn’t always been easy. At the time of the IPO, Grubhub was trying to get restaurants to move from the fax-based order flow to a tablet, and there was some strong pushback. The stock had a rough day when that was a headline for a little bit. Worst yet, Grubhub has made some enemies in the restaurant base, especially amongst restaurants who already had robust delivery businesses to start with. It was hard for them to stomach a 15% take rate on what they perceived as their own organic demand. The fact is people just like ordering online too much. One restaurant was quoted in the Tribeca Citizen as saying, “If I stop using them, tomorrow, I close the door.”
On the one hand, there’s no better summary in my mind of the moat of Grubhub. It is literally the source of demand for the restaurant. On the other hand, restaurants like that fear how much of their destiny is no longer in their own hands. Keep in mind that 90% of independent restaurants close within the first year, and the survivors average a five-year lifespan. It’s a cutthroat industry. It’s hard to parse out exactly to what extent Grubhub is merely a scapegoat for the inevitable failures or a rightful target of scorn for taking a big slice of margin in such a low-margin business. More recently, there’s been controversy about Grubhub’s system of protecting against those who work around the online order flow by calling in orders. The New York Post has really been hammering away at the story. The company has systematized a charge at restaurants based on the average order value for phone calls over a certain length. it has settled with some restaurants, but even the larger settlements don’t amount to too much dollar value. I’m not too concerned about this risk as it seems like more like anger finding a channel to vent than true investment risk.
Restaurants do love Grubhub. In fact , plenty of them have built their businesses on it. Before Grubhub came around, the only game in town for delivery in most cities was pizza or Chinese food. Grubhub has empowered just about every cuisine to participate in this important and growing market. Many restaurants advertise, so they were spending money on accessing customers anyway, and Grubhub is viewed by many as just another advertising channel. However, it’s not just any advertising channel because the best part is that the restaurant only pays out of its budget if an order is actually made. This helps solve the ROI and attribution problem for ad spend. A really important point is that restaurants will attest to the high incrementality of orders from Grubhub, both in terms of the diners they reach and the average order value. AOVs tend to be higher for the typical restaurant in the delivery channel than elsewhere.
It’s also important to note that phone orders are a messy process where mistakes can get made. Mistakes frustrate diners and have to be corrected at the restaurant’s expense. There’s a cost in both sides because you need a person dedicated to answering the phone, and the restaurant needs to foot the bill for fixing a mistake. Digitizing order flow drastically reduces the error rate. Fundamentally, diners want to order how they want to order, and whether restaurants like it or not, they need to facilitate that reality. Seth Priebatsch, the founder of LevelUp and now a key figure at Grubhub, called this the Faustian bargain. Lastly, delivery orders and takeaway have minimal front of house cost, and so there’s some margin offset to balance the cost of the take rate.
Grubhub now has another side to its business, and that is delivery. This is where we get into why Grubhub had that first post-hype sleeper phase. When it got into the delivery business in late 2015 by acquiring three regional delivery services, there was a lot of negativity from the investor base. This negativity was amplified when CEO Matt Maloney was quoted calling delivery itself a shitty business. Yet, there was a method to the madness and a degree of misinterpretation, but Maloney had a point. There’s little to no margin to be made in delivery itself. While pure marketplace orders are a two-sided marketplace, delivery is a three-sided marketplace where diners connect with restaurants with orders, and the delivery itself is handled by a driver. That’s even harder to pull off from a technological and a logistics perspective. That’s an important point there. The marketplace take rate is about 15%, while delivery is 30% with a $2 delivery fee on top.
What Maloney was getting at in calling it a shitty business is that delivery itself is merely a means to two specific ends. There are a lot of restaurants which didn’t want to hire staff to do delivery themselves, so this gave them a way to get into it without actually doing so. Secondly, there is more opportunity for hungry diners, so there will be more orders.
The key for the company became fitting the cost of delivery orders into the take rate that goes above and beyond the pure marketplace take rate. Think of the take rate on delivery orders as having two components to it – 15% for generating demand and 15% to cover the cost of delivering an order. Since orders on marketplace and delivery carry very different margin profiles, Grubhub introduced the concept of economic parity for investors to follow progress. What the company means when it mentions economic parity is that inevitably, if the cost of delivery is covered perfectly, that 15% piece of the total 30% pie, then the EBITDA per order should be the same between delivery and marketplace orders. This would leave every stakeholder in the platform agnostic to whether an order comes via the marketplace or delivery side of things. One important point to add here is that management has made a commitment to diners that any efficiencies on the delivery side which would conceivably result in margin on the 15% piece will be given back in the form of lower fees.
Grubhub has a much better margin on marketplace (which it calls self-delivery) than on Grubhub delivery, but the profit pool is the same. Grubhub’s profit per order in both cases is $3.40 based on a $30 average order size. Many investors focus intently on margins, and purely on that basis, delivery looks like a worse business, but in reality, the objective at Grubhub is to create orders. In that sense, the amount of gross food sales (GFS) and the EBITDA they generate is far more important to figuring out what this is worth long term.
Let’s now talk about making that shitty business yummy. Think of late 2015 through the end of 2017 as the first wave of Grubhub’s push into turnkey delivery. The company achieved economic parity, and the stock blasted off after getting comfortable with the transformed business model. This success built management’s confidence in the company’s ability to roll out new delivery markets and bring them to scale. It learned a lot of lessons along the way and wanted to take advantage of that. As a result, management decided to get even more aggressive. This came in the form of a partnership with Yum! Brands, with Yum! taking a $200 million equity stake. The partnership and proceeds from the Yum! stake were intended to help Grubhub support the cost of expanding massively beyond its existing market into new Tier 2 and Tier 3 cities.
Big chains come with great customer awareness and volume. This would help introduce the Grubhub brand to new diners. Yum! explicitly agreed to support this in its own marketing, and it did a great ad with the Taco Bell launch on Grubhub’s platform. Grubhub didn’t pay a dollar for the actual ad, but it played a prominent role. This goes above and beyond the organic awareness that being the go-to platform for these actual chains creates. Grubhub would be able to leverage Yum!’s strong brands as a toehold in each new market launch. Effectively, Yum! would provide a path to acquiring diners at a lower cost for Grubhub and drive order volume in order to ensure the third side of this three-sided delivery marketplace, the drivers, would have enough work to be efficient. For the roll-out into new Tier 2 and Tier 3 cities, Grubhub was spending an incremental $40 million in operations and support, which was a real drag on margins for this year.
For Yum!, the partnership was the best way to get into delivery. It did a lot of research in advance and chose this path as it was cheaper and more efficient than building its own network. It came with provably high order incrementality. It’s a meaningful point for all restaurants which go on Grubhub, and Yum! is good proof that this is the case. It would tap into a new source of demand in the form of Grubhub’s nearly 20 million active diners. With Grubhub, Yum! will get shared data on the end customer (something other platforms were not offering) and would have the opportunity to treat each of its brands as necessary. For example, Pizza Hut would enter only on the pure marketplace side, while KFC and Taco Bell will take advantage of Grub’s delivery offering. This wouldn’t be an option with competitors like DoorDash or Uber Eats.
Yum! would also receive another benefit, which has become a third pillar of Grubhub’s strategy. Shortly after the Yum! deal, Grubhub acquired LevelUp and brought its founder, Seth Priebatsch, into a key role as its head of enterprise sales. Internet he process, Grubhub became a fully integrated technology partner that builds the website and the app for restaurants, including two of the Yum! brands. Furthermore, while the marketplace had always handled takeaway, LevelUp made it clear that Grubhub’s essence, its reason for being, is to digitize order flow, not just facilitate delivery orders. This is important both in terms of how the product is built and how Grubhub differentiates its offerings from competitors on the restaurant side of things. With LevelUp’s technology and a focus on digitizing orders, the company would be able to help restaurants build out their own digital presence, and Grubhub could integrate restaurant reward programs into its platform and manage how it’s treated across channels.
I strongly recommend that anyone with even modest interest in the space watch Seth Priebatsch’s talk at ICR in 2018 on solving the digital channel dilemma. He gives some great context on how restaurants could use rewards and the data that comes with cross-channel and the advantages that come with full data and analytics and cross-channel awareness. I think it’s the future of the space.
Everything was going smoothly in the market on the heels of these actions (the push for EBITDA parity, the Yum! partnership, and LevelUp), and the stock was in a very bullish phase. In the fall of 2018, however, things changed. The rollout of Yum! started happening, the EBITDA per order started coming down, and the stock entered a prolonged second bust phase.
DoorDash is generally considered one of the biggest catalysts here. It started 2018 seemingly on the brink, but a series of four massive financings ever since and aggressive and strong execution from the company transformed it from a laggard to what some now consider the leader in the industry. DoorDash became one of the poster children of SoftBank’s aggressive investment in cash-burning startups with no mandate for profitability. DoorDash was able to use its blessing to burn cash in order to aggressively invest in customer acquisition. Every monthly market share report from the credit card and email panels at data aggregators started influencing the Grubhub stock because they were showing DoorDash increasingly taking share and perhaps exceeding Grubhub. Importantly, though, Grubhub’s market share in its most important markets never slipped too much. In its main market, New York, Gurbhub maintains a very high share, and DoorDash is not too significant there.
This gets at an important reality of delivery – the network effects are not national in nature. What matters is the scale in a given region or municipality. The second important point here is that not all delivery markets are created equal. Some are way better than others. New York City is easily the best delivery market in the entire country, with far higher order values and much greater diner frequency. Marketplace as a percent of orders is much higher, so actual turnkey delivery is lower. Thus, profit and margin per person are much greater. In other words, the LTV of each diner in New York is far greater than in any other city.
To add to those two points, market share itself is only half the story. First off, the actual market is growing rapidly such that even in losing share, Grubhub’s growth is incredibly impressive. Second, a lot of the incremental growth has been in Tier 2 and Tier 3 cities where DoorDash had a first-mover advantage over Grubhub. Core markets for Grubhub where diner LTVs have been higher have been very stable. New York, Chicago, and much of the Northeast are regions are far more suitable to delivery than some of the more rural, middle America regions, which is where DoorDash has made most of its ways.
Third, a lot of DoorDash’s gross food sales that credit card panels pick up include non-partner restaurants. Unlike Grubhub, which develops a relationship with every single restaurant on its platform, DoorDash will sell food from restaurants which have no nexus to the company, so they’re not partners. When an order comes in, someone at DoorDash places the order at a restaurant and a driver picks it up. By the way, that driver has to pretend to be the person who placed the order and has to effectively lie to the restaurant. The compensation to DoorDash from those orders comes from diner fees added on top of the menu prices, so there are significant markups to the consumer. Many restaurants hate this for having no control over the quality of the experience and no visibility into who the diner is. For diners themselves, the fees are exorbitant.
Fourth, DoorDash has a deal with Walmart for last-mile delivery that gets captured in gross food sales. That’s not even food. What it comes down to is you’re effectively comparing apples to oranges when you look at things like national market share and total food sales. One other point to make is that Grubhub is wildly profitable on all of its key unit economics, whether it be each delivery or the average LTV of a diner, whereas DoorDash is not.
In reality, the signs of competition are extrinsic, not intrinsic to the situation. Grubhub’s own metrics are doing just fine. You wouldn’t even know it was facing a different competitive environment today than at any point in its past public history. There’s no sign of competition showing up yet in the pace of new diner adds, the churn among diners, or the order frequency. By the way, order frequency is trending lower, but that’s because New York, as one of the first early markets, has really different frequency from the new Tier 2 and Tier 3 cities. No signs of competition are showing up in take rate or the growth trajectory of the marketplace or delivery itself.
The market is reacting in fears of DoorDash, but the fears aren’t showing up anywhere because Grubhub does have a differentiated offering. It is chugging along because the offering is unique. There’s good evidence of this in Amazon shutting down its food operation less than two years after entering the space. Clearly, capital alone is not enough to compete, nor is having a big trove of existing customers. I want to make the point that Grubhub’s marketplace is the single most unique asset in the entire industry. It gives the company the mandate to focus on digitizing order flow instead of just being a commoditized logistics solution. At the end of the day, that’s what turnkey delivery is. Takeaway could eventually be as important as delivery itself. One other point I’d add to this on the marketplace is that its lush cash flows help finance delivery, so the company doesn’t need to take aggressive financing from the SoftBanks of the world.
When people talk about the stock with dramatic headlines about delivery wars, remember that only a portion of Grubhub’s business today and in the future is delivery itself, and a lot of its addressable market exists in digitizing order flow. This idea of digitizing order flow gives Grubhub the opportunity to focus on lowering consumer-facing fees. It has the lowest fees in the industry, including when compared to the pizza companies like Domino’s. That’s important because consumers are highly sensitive to fees, and some surveys show clearly that one of the biggest impediments to increasing consumer engagement is high cost. Another important source of differentiation is that the chains get all the attention and are an especially high percent of DoorDash’s demand, but the fact is the majority of the orders and the profit pool in this industry exists in the incredibly fragmented long tail of independent restaurants. After all, most restaurants are these mom and pops.
Grubhub is effectively outsourced marketing technology, and in cases where it does the delivery, it’s the logistics for the mom and pops. One thing Grubhub learned with the Eat24 acquisition, where the promotional activity had been a big part of driving demand while it was under Yelp’s ownership, is that promotions designed to induce orders are worthless or even a net negative to LTVs. They don’t meaningfully change the long-term behavior of any cohort of customers, resulting instead in what management calls rented orders. These rented orders push volume through and accomplish little else. When you’re not at EBITDA or economic parity in a given region, there’s some benefit on efficiency for the driver side of the network, but in any other situation, there’s no point. As a result of the lessons from Eat24, Grubhub changed it promotional behavior across the entire platform to accommodate this reality. DoorDash happens to be the most promotional actor in the industry, so take that for what it’s worth on the competition front.
There’s an extra point on DoorDash worth mentioning. Its driver pay model is different from Grubhub’s and arguably illegal in many jurisdictions. That’s TBD, but I think we’re going to find out soon enough. DoorDash effectively skims tips from drivers in order to support its unit economics. When a DoorDash driver gets a tip on an order, the company pays that driver less than what they’re supposed to get out of the take rate or delivery fee. Were it forced to change this, and I think it will be, its unit economics would deteriorate meaningfully, and its losses accelerate considerably. Grubhub does not do this at all. A tip is a tip, and that’s what the diner expects too.
One of the benefits of being involved in a position long term is how you can simplify a thesis. Given that competition is extrinsic, not intrinsic, what is the driving force here? I think the fact that the competition story emerged alongside management’s expansion into new markets has caused some confusion. Management meaningfully took down EBITDA per order to facilitate the rollout to new markets. This made some investors nervous about whether it’s competition or strategy driving profitability. However, CFO Adam Dewitt has been insistent that were growth not a priority, the company could earn $2 in EBITDA per order right now versus the approximately $1.30 to $1.35 it will realize this year. The fact is the growth runway is so long that management is focusing on capturing the opportunity while staying disciplined and methodical about how that’s pursued. The core profitability shines through when you break down the numbers in more detail.
Let’s talk about valuation. Price-to-sales is one way I like to think about growth companies. This is about as cheap as Grubhub has been, though importantly, the nature of price-to-sales has changed over time with the take rate differential on marketplace versus delivery orders. I’m talking about the structurally lower margin in delivery than in marketplace, and that should have some effect on price-to-sales. We could think about how to quantify that. One important thing I like to do with price-to-sales is how to think about your implied margin and growth expectations. At roughly today’s price-to-sales, either the margin expectation or the growth expectation is too low. If you assume 5% growth in perpetuity, you get a 20% net margin, which is within the range this company was even in growth mode and will be down the line. However, it is growing top line, so it’s best to talk about normalized profitability in contrast to where it is with the delivery rollouts. Normalized profitability growth is upwards of 20%, and it should be in the double digits for some time into the future.
A second way I like thinking about valuation here is market cap to gross food sales. It’s my favorite valuation metric for the stock. In its public history, the company’s numbers fluctuated at around a 1:1 ratio. What’s important is that the market cap/GFS captures the compositional shift from marketplace to delivery orders that price/sale does not capture. Effectively, EBITDA per order becomes the implied margin here. The 1:1 ratio has persisted while EBITDA/GFS has hovered around 4.5%. In essence, the stock’s value should grow roughly in line with the growth in gross food sales, which increased 34% in 2018. We expect about 27% growth in GFS this year. Over the next five years, we expect GFS to compound at about 18.3%, with room for upside. If you think about 18.4% as an expected IRR on the stock, that’s not a bad return in any investment situation.
While EBITDA per order has been under pressure, we expect it to recover as the $40 million in delivery market expansion expense rolls off and new markets get more efficient. Further, we think this will naturally trend towards the $2 EBITDA per order Dewitt says he can get tomorrow and beyond as the company resumes demonstrating how much scalability and margin leverage there is in the cost structure. S&M, which is the company’s customer acquisition cost, is one of the biggest sources of operating leverage. As S&M’s percent of revenue declines, the company should be able to get to a 7% EBITDA/GFS margin down the line when you get closer to maturity.
Operations and support assumptions give you a sense for when and where we expect total efficiency on the delivery side where the cost structure fits into that 15% part. A big lever on the delivery side, in particular, comes in the form of average order value. The cost per drop, as they call in the industry the cost per delivery itself, is relatively fixed, so that high order value multiplied by the take rate flows through pretty purely to the bottom line. Given the really long runway the company’s growing into, there’s a good chance that the actuality of a terminal value is pushed out, and every year you add to this creates meaningful value.
One thing I always like asking is how I would know if we’re wrong. I think there are two key ways in which this would play out. First is the competition angle. If we see signs of increased churn in early cohorts, that would show up in a big slowdown in the pace of net new diner adds alongside consistently high spend on S&M. If new diner adds slow because S&M comes down, that’s okay. The company is telling us it can’t acquire diners. That might have consequences for growth, but it doesn’t mean the early cohorts are churning up. Realistically, for a competitor to induce churn in Grub’s customer base, it would have to meaningfully reduce a friction, like make the process much simpler or offer lower consumer-facing cost. That’s not happening anywhere. In fact, Grubhub has the lowest cost for consumers you can find in the industry.
Second, this could be wrong if management continues to burn margin on new delivery markets without ever reaching economic parity. It would show there’s something fundamentally wrong with the delivery market, and management isn’t quite as disciplined as it has been historically. What gives us comfort is that so far, there’s no evidence of the former and no reason to believe the latter. Management has been highly disciplined and rational. You see this most on the M&A front. It hasn’t paid the high valuations that exist in the startup space. It bought good assets at good prices at low implied values per diner, Eat24 being a good example of this, so I see no reason to believe why that would happen.
Another thing I like doing is reasoning by analogy, and I think a great analogy here is Booking, or the company formerly known as Priceline. In many ways, the rise of online food ordering has a parallel to the rise of OTAs. From my seat, Grub looks a lot like Booking while DoorDash looks a lot like Expedia. Sure, Expedia, in aggregate, does more gross bookings, but Booking is drastically more profitable because of a superior business model. It has the agency model versus Expedia’s merchant model. The key part of the analogy is that Booking deals with the really fragmented European hotel landscape while Expedia deals primarily with the large US chains. The economics of a marketplace are inherently inferior when catering to chains and hotels because the chains become a consolidated source of supply and have far more leverage over the platform. They will be no different in delivery. The chains get a lot of attention in the delivery headlines, but independent restaurants make up the vast majority of Grubhub demand, much like small hotels do for Booking.
One important difference in Grubhub’s favor in contrast to Booking is that in delivery, the order flow from search and discovery all the way through begins in-app. It’s Grubhub that owns the customer, whereas with Booking, its discovery still begins primarily at Google although the company is trying to change that with its new strategy. It’s still one step down the funnel, which means there’s some more margin going to different areas. With Grubhub, it is the starting point for its 20 million active diners.
Just for comparison’s sake, Grubhub today is at Booking’s 2006 revenue run rate. Booking has compounded top line at 24% over these last 12 years. With each year of terminal value extension adding significant value to Grubhub, you can appreciate why this investment opportunity is so exciting today. I’m not suggesting that Grubhub will compound at 24%, but it could fall decently short of that number and end up with an outstanding investment return for us. Long story short, we love Grubhub as a GARP investment with a wide moat and a differentiated offering in a large, growing addressable market.
The following are excerpts of the Q&A session with Elliot Turner:
Q: You mentioned that delivery is a local economies-of-scale game. Combined with the notion that the internet trends toward “winner takes all,” does it mean the future is Grubhub winning the profitable delivery networks? If that’s the case, does it then subsidize the less desirable markets to win those as well or does it optimize for profitability?
A: To an extent, I think it’s a little bit of a “winner takes all” in a given region. New York’s tipping towards what looks like an 80-20 balance, from Seamless web and Grubhub in the lead to everyone else. In aggregate, I don’t think it’s going to be “winner takes all” on a national level, and there will always be a reason for other offerings to exist in a given market. It will be hard for anyone to win everything.
Further, I think there will be consolidation in the industry. Grubhub has been a consolidator, making over a dozen acquisitions in its history. I do think there’s a big opportunity to consolidate, and Grubhub is attractive both as a consolidator and an acquisition target.
One interesting thing to consider is who Expedia’s CEO was in the early phase of the OTA industry, when the company was also rolling up, figuring out what offerings worked in the market, and scaling towards two main platforms. That was Dara Khosrowshahi, currently the CEO of Uber. Dara has said that he sees the potential for Uber Eats to be addressing a bigger market than Uber’s taxi offering. He has reflected, in some context, on the mistake of not having bought Booking when Priceline did. Expedia had an opportunity, and it didn’t appreciate the agency model at the time. I think Dara would recognize that in something like Grubhub, there’s a major opportunity to integrate the first-party and third-party offering. If Uber gets the marketplace and integrates it into Uber Eats, he’s got a lot more cash flow to subsidize the launch into other geographies and to maintain some less profitable markets.
Obviously, this is speculative, but in London, Uber started doing a marketplace offering. It would be a lot harder to get into that in the US. I don’t think we will have a winner-takes-all situation, but I believe there will be natural impetuses for consolidation as a lot more of the market moves online and the push for synergies leads to more profitability rather than growth itself being the primary driver.
I’d add one more thing. Amazon left the London marketplace in the fall of 2018, and it has just bought its way back into the UK with Deliveroo. Amazon also likes running the first-party/third-party marketplace. Look at what it’s doing with its actual goods. It’s very possible that something like Grubhub is appealing to Amazon because it recognizes the need to get into last mile as well and the way to do it.
Q: What do you make of the argument that independent mom-and-pop restaurants may lose market share over time to the startup kitchens being set up outside of a downtown area and solely focused on delivery? Would those use a platform like Grubhub or try to go direct?
A: That’s a great question. Cloud kitchens are getting a lot of interest. Travis Kalanick, the founder of Uber, has said that cloud kitchens are his next project, and he’s been working on that space. One of the ways to get efficiencies on the delivery side of things is to do what’s called batching orders. Let’s say an order is placed with a restaurant. The driver goes to that restaurant, picks up an order, and takes it straight to the diner. That’s slightly less efficient than if a driver could go to one restaurant, pick up five orders, and then go to five different houses. Cloud kitchens make a much cleaner opportunity to batch orders.
Munchery was a startup in New York City that tried being a cloud kitchen doing its own delivery, and the model simply didn’t work. It’s really expensive to run a delivery network, and having a lot of offerings from one kitchen wasn’t enough. You need a lot of diners interested in your offering and a lot of supply on the restaurant side. You need depth on both sides of the marketplace. Still, I do think cloud kitchens are going to proliferate because of these delivery offerings in given markets.
In the past, Grubhub invested in some of these and still has them. Deliveroo in London has invested in them pretty intensely. One of the things Grubhub does is use data to figure out which markets are missing an offering that the cohorts of diners tend to like. It could look at diner behavior and say, “In this market, people who like this pizzeria, that Chinese restaurant, and that Italian restaurant would like an Indian offering.” Then, it can tell the restaurants in that region there’s an opportunity, and the restaurants could use their back of house and create some sort of second restaurant out of the same address, where they don’t necessarily have that second restaurant out of the front, but they’re able to batch orders that wat. Long story short, I do think cloud kitchens are going to play a big role in this industry, but I view the platforms as the grease for the wheels. I don’t think they’re competitive. They actually play greatly into the advantages of the platforms.
Q: Is there more data available on the market share, also versus delivery alone, and in various local markets or nationally? Can you provide more details?
A: The marketplace is the biggest part of the offering, with about a third of the orders going through delivery itself. Because of the compositional shift from Tier 1 cities like New York and Chicago to Tier 2 and 3 cities, the market is going to be a lot bigger in different places, and we expect delivery to start taking more share. You could effectively back into it because credit card data gives you aggregated market shares, so there’s $20 billion flowing through the delivery platforms. You have 2/3 of Grubhub’s gross food sales flowing through marketplace. Divide that number by the $20 billion on platform TAM, and you will have the share that’s marketplace.
It’s important to point out that while the share going across marketplace is decreasing in Grubhub’s business, marketplace itself continues to grow at a nice rate. There will always be a place for it long term. More restaurants are starting to realize that as their delivery business does better and better, they have some margin to capture by having their own delivery offering. As this market shakes out, you will see a second wave where marketplace recoups some share total from delivery. We are not embedding that into any assumptions longer term, but I think marketplace will play a really important role and has the best economics long term.
Q: Regarding DoorDash, it sounds like the competition could be called irrational in the sense that DoorDash is losing money to try to gain market share. No market space seems great with such irrational competition. How do you see this playing out longer term? Is there a scenario where both DoorDash and Grubhub do well, that is, they both make money?
A: The market could stay irrational longer than you could stay solvent, right? There could always be follow-on cash behind the irrational cash, but to an extent, there is a rationale behind it, which is the massive opportunity. I believe there’s room for both DoorDash and Grubhub to succeed as they’re solving different problems. DoorDash is solving last-mile logistics, Grubhub is digitizing orders, and there are different ways that this will play out over the long run. Look at Uber IPO-ing now. Incidentally, SoftBank is in both Uber and DoorDash. As you go public, the market gets increasingly focused on what you can do from a bottom-line basis. DoorDash investors will want a return on their investment at some point, and that involves an IPO.
At the end of the day, what gives me comfort more than anything else is that Grubhub’s core markets remain as profitable as ever and continue to grow. Diner behavior continues to be consistent. Irrationality doesn’t change a solution which works comfortably, seamlessly, and consistently for diners. It’s mostly around the edges, and I do think there’ll be room for them both to succeed. It goes back to the general idea that this is a local business, not a national one, so what works in one geography won’t necessarily work in another.
About the instructor:
Elliot Turner is a co-founder and managing director at RGA Investment Advisors, LLC. RGA Investment Advisors runs a long-term, low turnover, growth at a reasonable price investment strategy seeking out global opportunities. Elliot focuses on discovering and analyzing long-term, high quality investment opportunities and strategic portfolio management. Prior to joining RGA, Elliot managed portfolios at at AustinWeston Asset Management LLC, Chimera Securities and T3 Capital. Elliot holds the Chartered Financial Analyst (CFA) designation as well as a Juris Doctor from Brooklyn Law School.. He also holds a Bachelor of Arts degree from Emory University where he double majored in Political Science and Philosophy.