Disclaimer: This interview is for informational purposes only and should not be relied upon as a basis for investment decisions. In Practise is an independent publisher and all opinions expressed by guests are solely their own opinions and do not reflect the opinion of In Practise.

{audio:0:00:20} Annanth, can you provide a short introduction to your background, please?

{audio:0:00:25} I joined Uber back in 2016. I was part of the US Canada operations team and was primarily focused on Canada, but spent a lot of time working with our US counterparts and some US specific projects, as well. When I joined, we were a pretty small and nascent team and Uber was mostly just in the Canada, US and France and so hadn’t really done its global expansion yet and we were just shutting down our original Uber Eats product, which I can touch on a little bit more later.

{audio:0:00:50} I originally joined to lead our sales and our onboarding team, mostly focused on acquiring and retaining restaurants. Then I spent a few years leading our business operations team, primarily focused on cross-functional initiatives that didn’t really constrain itself to a specific customer. For example, some of the initiatives we used to run were strategic planning for the business, managing and owning the P&L, launching new markets across the country, competitive intel, launching new products – we actually launched our corporate product and our grocery product, for example – and then business development, so working with partners such as McDonald’s and Yum! Brands, to get on the platform and to grow their business.

{audio:0:01:29} Could you lay out the Canadian food delivery market structure, before you joined?

{audio:0:01:36} It’s really interesting and I think Canada is in a really unique position. Many people don’t know this, but Uber Eats was actually launched in Toronto, back in 2015. The back story goes that Travis realized that, essentially, if we could get a car or a driver delivered to you in five minutes, what else could we get to you? We started this new product line, called Uber Everything and it was just a bunch of experiments. They originally picked Toronto because they didn’t think it would get picked up by the media very much if things went horribly wrong.

{audio:0:02:10} They didn’t; they launched a product and, in the first week, they did thousands of orders, so the signal was quite strong. Uber Eats ended up scaling from there, into the US, San Francisco, France and so on. Canada has a really unique position in the sense that it is Uber’s most mature market. But we were actually a late entrant into Canada and food delivery, in general. If you look at other competitors, such as Just Eat and Grubhub, they had already been operating for years. In terms of Canada, specifically, obviously we had Uber, we had Just Eat, who had a marketplace business, we had SkipTheDishes which was eventually acquired by Just Eat; we had DoorDash, which is also operating in the US and now in Australia. There was a company called Hurrier which was eventually acquired by Delivery Hero, which is a global food delivery player and they rebranded that to Foodora, which is their more upscale logistics business.

{audio:0:03:07} That was the context I entered in, back in 2016. Just Eat, historically, was by far the market leader, with their marketplace business. Eventually, I think they came to the conclusion that Canada was not a marketplace type of country, unlike the UK or certain markets such as New York. There just weren’t enough restaurants that did their own in-house delivery. Obviously, that’s problematic because you don’t really get that selection and that flywheel going. Uber scaled so quickly that they decided to fold their Just Eat business and just completely closed shop and they acquired SkipTheDishes back in 2016, right around when I joined. SkipTheDishes was a logistics first business; they didn’t really have a marketplace business. Essentially, they not only generated the lead or the order for the restaurant, but they also took care of the last-mile logistics, using their own driver network.

{audio:0:03:59} At the time, the industry was pretty nascent and was probably doing less than $500 million in GMV; it is now obviously doing several billion dollars and it’s been growing 100% to 200% per year, since 2015. Foodora has exited the market so the context has changed, over time. But essentially, that was the structure when I joined, in 2016.

{audio:0:04:19} It’s interesting, that point you made on Just Eat’s strategy of actually closing down the marketplace and saying, this is not a marketplace country. When looking at that kind of competition between marketplaces and logistics, what really defines a marketplace country?

{audio:0:04:38} There are a few countries in the world where, partly for density reasons and partly because of culture, restaurants historically have just not done their own delivery. Australia is another good example where it’s a very logistics heavy business. If you go to London, UK, for example, you will see a ton of restaurants that do their own delivery. It just has a long history, over the last decades, with Chinese restaurants, pizza restaurants, all doing their own delivery so it’s pretty historical. That is the underlying structure that really drives what type of business is going to be successful later on. You can’t really create a marketplace business out of thin air. If restaurants don’t do their own delivery, you’re not going to force or coax them into that; it’s really just pushing a boulder uphill. Most of these platforms have come to the realization that it’s better to tailor your business to the underlying structure as opposed to trying to change the underlying market itself.

{audio:0:05:35} Did Just Eat switch their supply side, in the marketplace, to SkipTheDishes?

{audio:0:05:41} They actually totally folded their original marketplace business. They shut down operations, the brand, everything gone. Employees were transferred over to the SkipTheDishes business. That business operated out of Winnipeg and it has always been a logistics business. They eventually ended up adding a marketplace business but, to this day, if you look at Canada and, to a certain extent, some markets such as Australia, it’s probably 90% to 95% logistics. Very few restaurants do their own delivery and, I would say, there is a secular trend, in these markets, away from doing their own delivery because logistics is such a turnkey service. Running your own in-house delivery is pretty complex. There are a lot of headaches related to it, such as hiring drivers, getting commercial insurance, paying them on time and so on. You are seeing a secular trend towards logistics business, around the globe.

{audio:0:06:32} On this point, do you think that’s going to really continue in those strongholds, like the UK or Netherlands, where historically, the marketplace business is? Do you think that we’re going to see the restaurants say, actually, it’s a bit of a headache doing delivery; I’m just going to pay the fee for Deliveroo or Uber Eats?

{audio:0:06:51} I think that’s the million-dollar question. Nobody actually knows what’s going to happen over the next couple of years. There are two things, I would say. If you look at Just Eat’s logistic business in the UK, for example – they just recently released their Q4 results – they had a stellar quarter for their logistics business; I think it grew about 300%, year over year. They’ve been doing a good job more recently, transitioning to a logistics business.

{audio:0:07:17} There is this underlying question around, if you do your own in-house delivery now, will you change to logistics? Part of it is the nuance of these restaurants. For example, if you’re a mom-and-pop restaurant and you don’t have any staff – you’re the cook, the chef, the manager – but you have some kids who are just always hanging around the restaurant, that isn’t an incremental cost to the restaurant, to do that delivery, to send your son or your daughter to go out and do that delivery. That’s not going to go away, per se. To a certain extent, you are always going to have a pretty sizeable chunk of restaurants doing their own in-house delivery, when somewhere is already a marketplace driven market, such as the UK, for example.

{audio:0:08:01} The million-dollar question is not necessarily what is going to happen over the next five or 10 years. It’s going to be, over the next 10 plus years, will those mom-and-pop restaurants, who leverage their own existing family members or cheap labor, still be around?

{audio:0:08:17} Going back to Canada, what was your strategy in entering, post 2016?

{audio:0:08:24} Scaling in Canada was interesting. When we first launched, Uber Eats was a pretty differentiated product versus other players. A lot of it was minor differences. For example, we had pictures of the restaurant and food. Believe it or not, that was not a thing before Uber launched; that was a small product innovation from us. We also launched with flat delivery fees; $3.99, no matter where you are in the city, you can get a restaurant delivery versus these really complex pricing structures that some folks had. No minimums. Grubhub, Just Eat, etc. usually had a $15 or $20 minimum.

{audio:0:08:59} We launched, originally, with a pretty differentiated product; we were pretty aggressive on pricing and promotions and we really focused on providing a better experience. Part of that just comes from the fact that we were integrated. For example, in a marketplace business, you cannot see where your driver is in terms of, are they picking up the food, how far they are. We brought over one of the innovations from the ride side of the business, where you can open up your app and you can see, at any time, where your driver is. He’s five minutes away; great, I’m commuting to home and I’m going to be there at the same time and I don’t need to call them to tell them to leave the food at the front door. It was a pretty differentiated product and experience and it had really strong product market fit and started to grow organically.

{audio:0:09:43} Our real bottleneck to growth, for the first couple of years, was actually just scaling the business geographically. The challenge we had is not necessarily unique to Uber; it’s a classic marketplace issue of the chicken and egg problem. You need customers for your supply and then you need supply for your customers. Uber was pretty lucky in the sense that we had a rides business, where we had an existing customer base we could tap into; we had a driver base that we could tap into. We spent the first couple of years, actually just acquiring restaurants. We started in dense urban cities, such as Toronto, Montreal and Ottawa and then we would spread out to the suburbs and, finally, we moved into tier two or tier three cities, so smaller cities.

{audio:0:10:26} After a couple of years, we got to about 80% population coverage, which was awesome. But then the growth started to entrench away from new geographies and more towards growing your existing markets. That was a totally different ballgame. We needed to add chain partnerships, like McDonald’s. We were actually one of the first platforms to sign with McDonald’s, back in 2017. We started to invest heavily in marketing; both brand marketing and performance marketing. We started to think about how to reduce the price so that we can drive more trial and retention for existing customers. You saw the prices go down, from $3.99, down to $2.99. Finally, we were really focused on removing defects in the system. When I say defects, I mean preventable issues for customers, such as giving your customer the wrong order; they ordered a burger but they got a hot dog. Or the driver was late by 20 minutes because they went to the wrong address.

{audio:0:11:25} That’s really helpful because it helps to increase your retention, which is another way to grow your business as opposed to the acquisition side. Those are the areas that we focused on, post 2017, after we had done all of our geographic launches and I would argue is still, to a certain extent, where a lot of the platforms spend their time.

{audio:0:11:43} How was the selection different? Taking SkipTheDishes, as the incumbent, how would you compare the selection on Uber Eats versus Skip?

{audio:0:11:53} Skip did a phenomenal job of building up a salesforce in 2015 and 2016 and they moved quite rapidly. I think they had a key insight which was, of all the axis of competition for platforms, it’s not necessarily speed or price or your product that is really the bottleneck to trial and retention; it’s actually selection. They did come to that realization faster and so they built an SMB and a enterprise team quite rapidly. From the early years, both from a quality and quantity perspective, they did an amazing job versus Uber. They had a first-mover advantage on the chain restaurants so they were able to acquire a lot of chain partnerships, largely through exclusivity, which created a differentiated product for them. They actually went out and leveraged that differentiated chain selection to both acquire and retain customers.

{audio:0:12:47} I would say, over the last couple of years, you have probably seen all of the platforms in Canada – and to a certain extent, in the US as well – move towards parity. Now chain partners are multi-platforming so they are with a lot of platforms. SMBs have also started to understand the value of being on multiple platforms. Selection used to be quite differentiated across platforms but that is less and the case nowadays, I would say.

{audio:0:13:11} Less and less the case mainly in the logistics markets?

{audio:0:13:15} Yes; and in marketplace, to a certain extent. Marketplace is interesting because there is a different sort of network effect there. Going back to my example of the Chinese restaurant who is employing their two kids, those two kids have limited capacity. If a marketplace business can actually fill up that capacity with enough orders, that restaurant is not going to add another platform. They are supposedly pretty happy with the orders and they can’t handle anything else. You don’t really see that dynamic on the logistics side so much because there is infinite supply on the driver side. You still have some differentiation on the supply side, on the restaurant side, for marketplace businesses but, again, that’s less true in the US and Canada, but it is very true for some marketplace heavy type of countries, like the UK.

{audio:0:13:59} The restaurant that is on the incumbent marketplace, they are less prone to testing the Ubers or the DoorDash because, as you said, they’ve got their kid who delivers and that suits them, for now. That’s the million-dollar question; could that change?

{audio:0:14:21} Exactly. Especially if restaurants are pretty happy with the sales volume that they are getting and they don’t want to add a lot of operational complexity. It is a pain in the butt to have two, three, four tablets going off all the time and if you’re not getting enough incremental orders from all those tablets, a lot of restaurants will just say, hey, I’m happy with this one platform.

{audio:0:14:41} The second way restaurants sometimes get locked in is actually through explicit exclusivity agreements. Again, that is a lot less ubiquitous nowadays, as restaurants start to realize that you can drive a lot of incremental orders from a second or third platform. But platforms are savvy enough to know that, if you get certain key restaurants in a market or in a city, and you can get them exclusive, you might be able to leverage them to the point where you can create a really differentiated product and that alone will allow you to acquire and retain users.

{audio:0:15:12} As Uber, you enter in Canada and you quickly start onboarding the supply side and then the suppliers realize, this is incremental volume and I can go on Skip and Uber, so then we see a selection converge across platforms. What is the endgame here, in logistics markets? What actually drives a competitive advantage?

{audio:0:15:37} That’s a good question. I think you see this across all sorts of technology and marketplace businesses. In the beginning, you have table stakes and so I think, for marketplaces businesses, like food delivery, that is going to be price, selection, speed, product and customer experience and across all of those axis, all of the players have, essentially, converged. For example, SkipTheDishes used to have about a 45-minute delivery time; Uber probably close to 30. Skip is probably close to 30 minutes, right now, after a couple of years of technological and operational legwork. The same thing on selection; close to parity. Same thing on price; close to parity, and so on.

{audio:0:16:16} The axis of competition that used to be differentiated have now become a bit commoditized. The market is pretty dynamic and has changed over time. To your original question, going forward, what’s going to be the point of differentiation; nobody has really cracked that nut yet. There are a couple of key areas and one of them would probably be continuing to invest in branded marketing. At the end of the day, these are consumer brands and much like a Peloton or a Nike, you can create a really strong brand and an affinity with your consumers where it gives you a slight edge over some other platforms.

{audio:0:16:51} The second area is probably going to be on product innovation. DoorDash has done a really awesome job of this in the last couple of years. For example, they have DashPass which is a subscription product; you pay a flat fee of about $9.99 per month and you have a zero-dollar delivery fee and you get a discount off your service fee. That’s a good example of a product innovation where it’s actually driving acquisition and retention. Very similar to Amazon Prime, where it’s changed the trajectory of the business and is helping to drive some sort of competitive differentiation.

{audio:0:17:20} The third place is new adjacent categories. This is probably the area that makes the most headlines. Can platforms extend beyond just food delivery and can they move into other verticals or products that are incremental, such as white label logistics for restaurants or grocery or convenience or pharmacy and so on? To a certain extent, they are copying the Meituan playbook in China, which is a super app strategy, where you layer on all of these additional verticals and each of them allows for a really cheap cost of acquisition and, hopefully, that bundling creates higher retention. Particularly in the US, you are seeing DoorDash move very quickly on convenience. The latest data I’ve seen is that they have, essentially, become the market leader, over the course of a year, which is awesome for them. You’re not really seeing that as much in Canada; I think the US is just much bigger and so that is where most of the focus is right now. But I suspect those are the three areas where you will start to see competition shift as people get to parity on some of the table stakes stuff.

{audio:0:18:26} If you just play that forward five, six, seven years and these logistics networks are scaled and have much higher volume and density, doesn’t that then become even more attractive for those marketplace restaurants to enter because the fee could be cheaper because you have higher density, for example?

{audio:0:18:44} It could be. What’s interesting is, globally, most of the marketplace businesses tend to be very focused on, firstly, defending their core business, because that business is being attacked from all sides. From an investor perspective, you’re seeing this interesting dynamic where these businesses had a great trajectory, they were the cash cows; Grubhub was churning out hundreds of millions of EBITDA per year, for example. All of a sudden, you have Uber and DoorDash who are well-capitalized, come in and, essentially, just poison the well and destroy economics in their core business.

{audio:0:19:16} They’re very focused on just trying to maintain their core marketplace business, which is a great cash cow. Secondly, they’re really focused on trying to develop a core competency in technology and logistics. Traditionally, they have been very good at sales, marketing and finance so building up these new muscles around logistics has been quite tough for them, I would say. I don’t think they are there yet, in terms of thinking about grocery, convenience, etc. The CEO of Just Eat takeaway has explicitly said that they will consider grocery but they are not looking into it. Grubhub has dabbled a little bit, but not really. They haven’t made any major moves in convenience, for example; they don’t do pharmacy, at all.

{audio:0:19:55} I think you may see marketplace businesses dabble in those areas in the future, but for the next couple of years, I would guess they are just super focused on the food delivery space itself.

{audio:0:20:07} But then the logistics players –Ubers, DoorDash, etc – when they do get to scale and they have high volume through the network, it’s going to become quite a risk again for the marketplaces, potentially, if they have that scaled network, in those suburban areas? For example, of the UK, where Just Eat have a stronghold, if Uber is at scale, then the restaurants are going to be saying, I might as well go through you at a much lower take rate?

{audio:0:20:36} I think maybe not from the restaurant’s perspective. I think the incremental growth, when we talk about these secondary verticals, is going to be, how do I get a consumer who uses and opens the app, two or three times per week, to then open the app four or five times? That fourth or fifth time might not be ordering from a restaurant; it might be ordering grocery or pharmacy. It might not drive incremental scale on the restaurant side, but it does drive incremental scale for the platforms, because they are just adding new businesses.

{audio:0:21:08} Where you do get really good scale is on the driver’s side. Drivers don’t care if they are picking up from a restaurant or a pharmacy and, all of a sudden, you add on all these businesses and those are incremental to the driver’s side and so you should get better operating leverage out of your infrastructure, your driver acquisition and so on, which should translate into lower prices on the driver’s side, over time.

{audio:0:21:30} That’s why it’s interesting because, potentially, in the long run, these two businesses can operate in parallel. The logistics players are building up a true logistics infrastructure, whereas Grubhub or Just Eat are, as you said, protecting the marketplace but also building the logistics to protect that.

{audio:0:21:53} Yes, exactly. To be frank with you, I think the report of Grubhub or Just Eat dying is grossly overstated. I think the best analogy I’ve heard is Amazon versus eBay. eBay is an amazing business that generates a ton of cash. Are they as big as Amazon? No, not really. But it’s still an amazing business. I think Grubhub, Just Eat and so on, those guys will do fine. There’s a lot of secular tailwinds to the industry and the penetration is quite low. Those will be great businesses one day; they’re not going away. The real question is, will they have the appetite and the capability to go beyond logistics and into these other verticals? For the time being, it doesn’t look as if they are actually that interested in them.

{audio:0:22:39} Going back to Canada, you mentioned how the selection is converging and it’s around the edges where you have an advantage. Was Skip mainly profitable in Canada, versus DoorDash and Uber who were still burning cash to enter?

{audio:0:22:57} Yes; I would say that Skip has, traditionally, been quite focused on unit economics and profitability. One of the advantages that they had was that they tended to focus on more suburban markets. You see a very similar dynamic in the US, with DoorDash. Suburban markets are great for unit economics for two reasons. Firstly, you tend to acquire more family type customers and families probably have a little bit less order frequency, but their average order value, their basket size, tends to be much bigger. It just makes sense; you are probably ordering for three or four people versus a young professional, downtown, that might be a single order. Basket sizes are one of the most important drivers of unit economics because out of every incremental dollar, anywhere from 30 to 40 cents goes directly to the bottom line, which is great, so it’s very accretive.

{audio:0:23:49} The second thing about the suburbs is that delivery drivers tend to just cost a lot less. One, delivery drivers tend to live in the suburbs and not downtown, so it’s a lot more convenient to work out of the suburbs; you don’t have to pay them as much to coax them to go into downtown traffic. The second thing is, when you look at the downtown core, the process of parking a car, picking up the food and then dropping it off, maybe being stuck in traffic, that requires a lot of time and energy and a lot of frustration on the driver’s side. On a per delivery basis, you actually have to pay a lot more. They are less efficient and they are a bit more frustrated than those in the suburbs where you are usually dropping off and picking up in a dedicated parking spot, like someone’s driveway or a parking lot. You get these really great efficiencies in the suburbs where you are making more revenue, you’re paying less on the driver side and that turns into much better unit economics. Skip really leveraged that, for the first couple of years and they actually just created a bunch of profit pools in their more mature markets, but they ended up leveraging that into downtown urban cores and using that to subsidize those business. But yes, they were much more profitable from the get go.

{audio:0:25:02} Can we just walk through the unit economics of a typical order? Maybe we could compare the suburbs and the city, as we go? Using rough numbers – and please correct me – I’ve got a food cost, mainly from DoorDash’s S-1, of $22, plus tax and payment fees is $3, so you get a $25 order value. Is that about right? How does that differ between the suburbs and the cities?

{audio:0:25:34} Give or take, that’s correct from most platforms; it might be a couple of dollars off. As I said before, I think, in the suburbs, you’re probably going to have anywhere from a $5 to $10 difference, in terms of order value – it really depends on the particular market – versus downtown core where you have more young professionals and single orders. That is the biggest driver in terms of order value. But that sounds about right for DoorDash, or the industry as a whole for US and Canada, at least.

{audio:0:26:02} How do you look at setting the consumer delivery fee for an order?

{audio:0:26:08} Over time, you can see the delivery fee has gone down. I would say, typically, most platforms are priced anywhere between $2 to $4; it really depends on the sort of level of subsidy and promotion. Consumer fees are made up of several components and, again, it depends on the platform. You have an underlying delivery fee that is anywhere from zero dollars up to $3.99 or $4.99. You have a service fee which is, typically, 10% or 15% and so that’s on the pre-tax or post-tax basket size. Then some platforms also have a small order fee. If your order value doesn’t meet a certain threshold, they charge you $2 or $3 to help cover the cost of the delivery.

{audio:0:26:49} So, it depends, but your all-in fees are typically going to range somewhere between $3 and $5, depending on the platform, depending on the market. If someone wants to be a bit more aggressive, to take share, they might discount that quite a bit. But that’s typically going to be your average delivery fee, in the US and Canada.

{audio:0:27:04} Then there is a tip on top of that that is required, from the customer?

{audio:0:27:09} Tipping is interesting. Uber ride-hailing, to a certain extent, really changed the perception of tipping in the gig economy. If you recall, for years and years, Uber did not have a tipping option. Compared to traditional in-house delivery, from a pizza joint or a Chinese restaurant, tipping is not as common on these third-party delivery platforms. The tipping rate can vary quite a bit; anywhere from 20% to 40%, 50% and I’ve seen some platforms that are very aggressive with their tipping features, which is up to 80%. Your typical tip is probably anywhere from $2 to $3. That is generally a pass-through, so it goes directly from the consumer, to the driver and the driver keeps 100% of that tip and the platform doesn’t take any of that at all.

{audio:0:27:55} Does the logistics model only work in tipping cultures?

{audio:0:28:00} I would say, not necessarily. Tipping is much more important to a marketplace and tip rates and tip amounts are much higher in a marketplace. That’s actually one of the reasons why marketplace businesses are a bit more profitable. Part of it is just the history and the culture there; part of it is just the type of consumer that uses the marketplace versus the logistics is different. If you look at DoorDash’s S-1, for example, and if you look at comments from Uber Eats, who have mentioned that they are profitable in quite a few markets, those are logistics markets. I think you are starting to see some early evidence or supporting data points, that the logistics business can be profitable. They don’t have to charge $8.99 delivery fees to get there. You just need to be really tight with your operations and cost control. You can have $3.99 delivery fees, for example, and still run a profitable business on the logistics side.

{audio:0:28:52} If you look at the actual fee to the driver, it obviously depends on the fee they get, per order or per hour, but let’s say it’s a $3 tip, on average, 10% of the actual order value. For the earnings of the driver, if they do one order per hour, then they need to receive another $5, $6, $7, from the gross merchandising value?

{audio:0:29:18} Yes, for sure. Typically, most platforms in the US and Canada are going to pay anywhere from $5 to $7, per delivery, and that’s coming directly out of the platforms; that’s excluding the tip. What you end up with is a $7 to $9 per delivery payout to the driver, before any of their costs or anything like that. Most drivers, in a more mature market, are probably doing about two deliveries per hour. It wouldn’t be unreasonable to expect $15 per hour, roughly, for a more mature market.

{audio:0:29:59} In markets where you don’t really have that scale, because you have just launched or it’s a smaller town or something like that, there’s a bit of subsidy that goes on, on the driver’s side. You might set an hourly rate and if a driver only does one delivery, the platform will make them whole by paying out the difference, for example. That’s really expensive and one of the reasons why platforms have been spending so much money over the last couple of years is because there is a certain level of subsidization that needs to happen as they are scaling the business. But I wouldn’t consider that a permanent dynamic in the industry. As these businesses scale, you should expect that to go away.

{audio:0:30:36} I guess, again, it all comes down to volume? If the driver is doing two deliveries an hour, you can make it work?

{audio:0:30:46} Totally. It’s a very scalable business. If you do not hit huge scale on the business, the dynamics of the marketplace just fall apart. Remember, drivers can come on and off the platform so you can have a ton of drivers just on the platform and you can’t really control that. If you don’t do enough scale, restaurants just churn out. They say, hey, why am I keeping this tablet on and trying to keep track of these things that only happen once every couple of hours. I’d rather just turn this off. The flywheel goes in the opposite direction if you don’t hit the right scale and that’s one of the reasons why there has been a bit of a land grab over the past of couple of years. All of the platforms realize that and that’s also one of the reasons why you’ve seen consolidation. Uber recently acquired Postmates; DoorDash acquired Caviar, for example. Those businesses hit scale in certain markets – Postmates in LA, for example – but everywhere else they were just burning a lot of cash. To your point, if you’re only doing one delivery per hour, that’s just not going to work for drivers.

{audio:0:31:50} That’s also why it’s interesting because, let’s say you pay $5 driver fee, per order, you’re still getting $3, $3.50 tip, so that tip is still pretty crucial?

{audio:0:32:05} Yes, it is important. If you look at the product evolution of these platforms, over time, they’ve been building in more robust and nuanced ways to get people to tip. It is an important part of their compensation but the platforms also know that they can’t be too aggressive with tipping or you will see some sort of erosion on your conversion and retention rates, so there’s a fine balance that you will see in the product. But there has been a lot of investment in trying to get people to tip.

{audio:0:32:33} The reason why I’m asking is because London is not as much of a tipping culture as the US and Deliveroo and Uber are fighting in London, versus Just Eat. Obviously, the density is going to be quite high, in London. But if you don’t have that tip, it’s just going to cost more for the platform, so the unit economics are going to be much, much worse than they would be for DoorDash in the suburbs, that are getting a tip of $3 an order.

{audio:0:33:00} Totally. I think that’s part of the advantage that Just Eat actually has in Europe. Their marketplace business is much more profitable and tipping is not as ubiquitous in the culture and so you have that bigger differentiation between logistics and marketplace businesses, along with higher labor costs, just in general, in Europe. That is actually a little bit problematic for these logistics platforms and I still think that they’re trying to figure out what the right balance is between supply and demand and earnings and trying to keep things affordable for consumers.

{audio:0:33:33} Then you’ve got Jitse going in and charging nothing.

{audio:0:33:40} Exactly. Essentially, a few weeks ago, the CEO of Just Eat Takeaway dropped delivery fees down to zero and I’ve seen some estimates that that might cost anywhere from £700 million to £800 million, per year, which is quite a bit of money for someone like Deliveroo, at least, to match. They’ve got a really interesting, competitive dynamic going on there. I would say it’s probably the most interesting competitive landscape in the world, right now. I think the US, in terms of DoorDash versus Uber versus Grub, has largely stabilized, particularly with DoorDash IPOing; investors are going to be demanding more proof of profitability from that business, so you will see a rationalization of costs, sales and marketing promotions and so on. You’re seeing the opposite in the UK where you are seeing a heightened level of competition with Just Eat, so I think that’s going to be a really interesting battle that goes on for the next one or two years.

{audio:0:34:30} Just going back to the unit economics, how do you look at optimizing the take rate?

{audio:0:34:39} Right now, the standard commission, if you want to call it that – or the marketplace fee – that is charged to restaurants for the delivery is about 30%; I’m just talking about logistics right now. For pure marketplace businesses, where you just generate the order and that’s it and you don’t do the fulfilment, you’ll see a bigger range there. Typically, it’s 15% but it goes as low as 10% and as high as 20%, in some mature markets.

{audio:0:35:03} There is probably only one way for a take rate to go up in the short term – the next one to two years – and it’s probably through marketing services that you sell to restaurants. For example, Uber recently launched a cost-per-click ad-based infrastructure technology so restaurants can log into the platform, they can set a budget and they can say, this is what I want to spend, and Uber will, essentially, boost that restaurant to the top of the feed, to give it more exposure. If you look at other markets like China, for example, they generate an enormous amount of their revenue just from marketing services.

{audio:0:35:43} I think you will have two types of restaurants take advantage of marketing services. One is chain restaurants, such as McDonald’s, who are folks who have a long history of just spending a lot on marketing and so they are quite comfortable with doing that. The second one is going to be SMB restaurants who have great margins. Those folks are going to have a high willingness to spend even more if they have the margin because every incremental order is going to go directly to the bottom line, even if the end margin is a little bit worse. I think that’s going to be a great business because, similar to Amazon, you’re going to be able to close the loop on the ad spend. I know that every dollar I spend, directly generated X dollars. If you look at something more top of the funnel, like Google or Facebook, you can’t always close that loop. Generally, I think restaurants are actually going to be pretty excited by marketing services and you should see that as a major growth area, over the next couple of years, for the entire industry.

{audio:0:36:39} As you said, it’s like Amazon’s inventory; it’s prime advertising real estate?

{audio:0:36:44} Yes; exactly. People really do value the ability to attribute a dollar of marketing to X dollars of sales. The early signals that Uber has publicly stated as the return on ad spend is actually quite good. I would expect them to continue to double down and focus on that part of the business.

{audio:0:37:06} Going down that income statement, you’ve got the driver costs that you mentioned, about $4 to $5 for the fee plus a tip, so let’s say they’re getting paid $8 or $9; plus you pay the merchant. I guess the next section on that is the refunds and promotions to drive that original revenue; how did that change as you scaled in Canada?

{audio:0:37:34} Generally what you are seeing is a rationalization or pull back on sales and marketing and promotion spend. Probably more so on the latter as the market matures and there is greater sentiment from investors as these platforms need to show some path to profitability. I think, for the next one or two years, and even in the last one or two years, you’ve been seeing heightened sales and marketing spend. DoorDash did a huge marketing campaign during the Super Bowl, for example. But promotion spend is quite large in the food delivery business; it’s very similar to ride-hailing. It’s very price and promotion driven, right now. That will eventually go away.

{audio:0:38:15} I think the other major line items are related to COGS and refunds. COGS are things like cloud infrastructure, driver insurance and so on. That’s a pretty material line item. Refunds are also actually quite a big issue in the food delivery business. The operational complexity creates a lot of opportunities for something to go wrong. At the restaurant, maybe they forget to pack your fries with your combo or maybe you order chicken salad and they gave you grilled when you really wanted something breaded. There are just a lot of nuances and people can be pretty specific about their food. Even on the delivery leg, maybe the driver gets stuck in traffic and so now your fries are cold and soggy. Refunds is a pretty significant line item for platforms, but they’ve been a good job of scaling that out, creating self-serve products and even just preventing them.

{audio:0:39:06} Would it be around 2% of GMV? 2% to 3%, for refunds?

{audio:0:39:13} That sounds about right. But again, you will see that come down. I think the platforms have been so focused on acquisition that they still have a bit of a leaky bucket and the operational experience is not 100% there. It’s much like Amazon. If you remember Amazon from the early days, there were some kinks in the system but, nowadays, your experience is, generally, pretty seamless. I would expect that to improve over the next couple of years and for that part of the line item to improve. A lot of that will be technology driven, through self-serve or other more sophisticated ways to prevent things from going wrong.

{audio:0:39:44} How did you see EBITDA per order or how would you look at calculating that for a logistics business?

{audio:0:39:53} It’s hard to predict only because no one has actually got to maturity and scale yet; the business is still so nascent and young. I would suspect that, in the long term, you will probably end up with 10% or 20% EBITDA margin. That will vary by platform. As I mentioned, DoorDash is a bit more suburban driven and so they will probably have slightly better margins than an Uber who is more urban and young professional driven. But I think 10% to 20% EBITDA margins is probably where you will land, in the long term and that will probably get you to $1, $1.50 in absolute EBITDA per trip. But we are a couple of years away from that, at least in the US. You still need to see a lot of rationalization on promotion spend, sales and marketing, scaling up the driver supply side, so you don’t need to subsidize as much and you need to see take rates go up, so you need to see the discounts end for large chain partnerships and you need to see the scaling of the ad business. When all of those things come together, you will probably see a pretty good margin expansion, in a couple of years.

{audio:0:40:52} What could break the economics, for logistics companies?

{audio:0:40:55} I think one of the bear cases for food delivery is, is it a defensible business or not? Do you just end up in this really protracted subsidy war over years and years and years, if not decades? I think that seems unlikely if all of the players are public. At some point, investors are going to demand a return on their capital. But the bear case, in the short term, is really a protracted war that, essentially, just destroys your unit economics and these businesses don’t end up printing any cash.

{audio:0:41:27} The other issue might be wholesale supplier pricing power. For example, McDonald’s, Chipotle, Wendy’s and so on, these large chains know that they drive a material amount of customer acquisition and retention for the platforms so they have pretty good bargaining power over them. There is this concern around the discounted take rates or commissions that they are paying right now. Is that going to rationalize, over time, or are they always going to have this power over the platforms? I think Spotify is probably a really good analogy here, with their music labels; very similar bear case there. Do these music labels constantly just drive a hard bargain and ask for a higher take rate? Or eventually, does Spotify have enough bargaining power that they can push back and negotiate for something better?

{audio:0:42:14} But we also had this issue of ride-hailing in the past.

{audio:0:42:20} I think ride-hailing is interesting. You’ve seen a consolidation of businesses. You don’t have the same wholesale transfer pricing because drivers are, obviously, quite commoditized. That’s one of the different nuances versus food delivery. You don’t really have that dynamic. But the good news on the ride-hailing side is, you did see rationalization on sales, marketing and promotion. The general consensus is that food delivery will follow a very similar path but it will take a couple of years; those things don’t happen overnight. There’s a bit of a game theory that goes on where you don’t want to rationalize too fast or else your competitors will see that and they will try and make proactive moves to take share. It tends to happen step by step, over a couple of years, at least.

{audio:0:43:05} This is also the difference in that the logistics businesses are typically in countries where restaurants don’t deliver and, therefore, there are more larger chains, such as McDonald’s and Wendy’s, which then have much more bargaining power than the likes of Just Eat or Grub, that seem to be delivering from the independent restaurants, which don’t have that issue.

{audio:0:43:25} Exactly. Believe it or not, McDonald’s is probably of the largest food delivery services in the world, right now. They’re probably catching up to Domino’s quite quickly. These guys are doing really serious numbers on the platform and it’s enough of a differentiator that, if you don’t have McDonald’s on your platform, you might actually just lose a huge chunk of your customer base to another platform. McDonald’s knows that; and so do all the other chains, such as Yum!, Wendy’s and so on. They have quite a bit of leverage in logistics-heavy markets or in markets where they are being used as a spearhead to attack marketplace businesses, like in London, for example.

{audio:0:44:01} Can McDonald’s really just pay zero? What is the end game in that relationship?

{audio:0:44:10} These platforms tend to view these chain partnerships as loss leaders. They are, essentially, customer acquisition chains. Instead of spending $100 million or $200 million on a marketing or a brand campaign, I might spend $100 million or $200 million on discounting a particular chain, for example. The justification there is that I get to own the customer relationship, I can cross-sell or upsell that customer into other services, higher margin restaurants and so on and so the LTV is still there.

{audio:0:44:40} Again, it’s unknown what will happen over the next couple of years. Will the platforms have enough leverage to actually increase pricing over these chain restaurants? I would guess that it is actually different, based on geography. For example, in Canada, there is some of the highest chain penetration in the world, in terms of restaurants; something like 52% of restaurants are chain partners. If you go to other countries, chains are not as ubiquitous, especially because chains tend to be franchise driven, so each country operates independently from each other, versus globally. You might end up with different pricing or different bargaining power in different countries so it’s not going to be a single solution across the world. You will see the dynamic play out differently, depending on the market.

{audio:0:45:25} It’s almost as if the logistics players need to use McDonald’s and the big chains to get them to scale, to then have a somewhat credible business, eventually, to actually ship anything; as you said, the groceries, the pharmacy. It’s almost like this is an interim stage.

{audio:0:45:42} Yes. When McDonald’s originally launched, back in 2017, with Uber, it was, essentially, the first chain to do that, across the world. McDonald’s did do their own delivery in Asia, for some time, and that was done in-house but really, it was the first chain to do that, at scale and the trajectory, with the industry, just totally changed overnight. DoorDash then went ahead and signed up Wendy’s, Chipotle and some other partnerships and so that set off a bit of an arms race, to sign up all these partnerships. But it was probably the most pivotal thing to happen in the food delivery business, in the last five years. It totally changed the dynamic and it changed the scale of the customers that you could acquire so it’s been a really important milestone, I would say.

{audio:0:46:25} McDonald’s is also on Just Eat, I believe, especially in the UK. Aren’t they going to be on everyone, eventually?

{audio:0:46:32} That’s a more recent development. Chains have become table stakes and chains know that and that’s part of the reason why there is this question mark around, will they continue to have pricing power? If Just Eat and Uber are going head-to-head, aggressively in the UK, and McDonald’s is considered table stakes, well guess who has got all the chips and the bargaining power? It’s McDonald’s. It’s not totally clear how that competitive dynamic will play out and who will be left with all of the power. But McDonald’s is definitely a savvy negotiator and they’re quite aware of the position out there and they are leveraging that quite well, I would say.

{audio:0:47:11} Just going back to the EBITDA per order, is that $1, $1.50 based on a minimum of orders per hour?

{audio:0:47:26} I would say, generally, you would probably need somewhere around two orders per driver to get there. In your smaller towns, for example, if you are doing 1 to 1.5 deliveries, you may not see those same margins. Though on the other end, you might see larger basket sizes, for example.

{audio:0:47:46} On the flip side, in more dense, urban cores, like where Uber operates, you might see higher driver delivery efficiencies, but you might see smaller basket sizes. There is a bit of a variability, across geographies, but in general, you probably need about two deliveries per hour to make the unit economics work.

{audio:0:48:04} Looking at this competition between marketplaces and logistics players, what is really the key question to be asking when analyzing who will win between incumbent marketplaces and logistics entrants?

{audio:0:48:23} That’s an interesting question because initially, when logistics players launched in the US, Canada and the UK, marketplace businesses were pretty dismissive of the model. The core issue they had was the assumption that it could never be profitable. The capital requirements to build out a driver network are quite immense and it takes a lot of technological and operational know-how or competency to build these out. Marketplace businesses have, traditionally, been super focused on sales and marketing; that’s their core competency and they’re quite good at that. Adding a new competency around technology and operations is actually quite difficult, if you’re not used to that.

Uber Eats, SkipTheDishes & Food Delivery Business Models

Head of Business Operations at Uber Eats Canada

inpractise.com/articles/ubereats-canada-logistics

Why is this interview interesting?

Executive Bio

Annanth Aravinthan

Head of Business Operations at Uber Eats Canada

Interview Transcript

Related Content

Online Food Delivery: Marketplaces & Logistics Business Models

Former CFO at Deliveroo & Finance Director, EMEA, Amazon

Deliveroo & Dark Kitchens

Former Head of Operations, Deliveroo Editions, Middle East

Scaling Uber EATS in London

Former Head of Uber Eats, London

Restaurant Franchisees & Online Food Delivery

Former Senior Director, Business Development at Subway

Audio

Uber Eats, SkipTheDishes & Food Delivery Business Models(February 9, 2021)

PARTNER

Speak to Executive