Former CFO at Deliveroo & Finance Director, EMEA, Amazon
Philip Green spent almost eight years at Amazon, culminating with the role of Finance Director of EU operations. He then took on the role of CFO at Groupon, followed by the role of CFO of Deliveroo. He currently is Director and CFO of theatre and entertainment producer Jamie Hendry Productions, as well as CFO of robotics and AR gaming business Reach Robotics, and advisor to several high-tech digital start-ups. Read moreView Profile Page
Phil, can you start by providing some context to the structure of the UK food delivery marketplace when you joined the market?
It was 2016 when I first got involved and I’d say that the UK market was very underdeveloped. You had significant success with Just Eat building a marketplace and then you had a number of early stage players in the market. Hungryhouse was still around at that time and restaurants hadn’t quite got used to the idea of doing takeaway food or home delivery. It was either you’re a takeaway business or a restaurant, so Just Eat was a marketplace for takeaway restaurants. Where Deliveroo came into the market was down to Will Shu as the founder, his experience of living in New York and having great access to good quality food, working late in the office and then moving to London and not having the same access to good quality food. So it was still very early on. Deliveroo at that stage was the leading player in the UK in the delivery market and was only doing £15 million of revenue, so still in the very early stages of that business. If you compare to today, Deliveroo has had to build a lot of infrastructure. People didn’t consider a delivery rider as a job. The gig economy was still relatively new. Uber was playing in the taxi space but the concept of the gig economy and the conversation around it was still very new.
What was the strategy of Uber Eats and Deliveroo, the logistics-based models, to come and attack that fat margin business of Just Eat’s marketplace?
Deliveroo’s starting point strategy was a highly curated selection, hyper local – it started in Chelsea – giving people access to good quality food in affluent areas. That’s where Deliveroo started. So a very selective, affluent customer base and also people who wanted good quality food, so attacking the higher end of the market. Then you had Just Eat entering the market in the UK in July 2016. When they entered, the game changed rapidly. Deliveroo was steady and successful. With Uber Eats entering the market too, all of a sudden you had this highly capitalized, aggressive company entering this space and changing the landscape, by very quickly increasing pay for riders. Their approach was very different; let anybody on the platform. So no real curation and as quickly as possible grow selection and try and own the rider supply chain, focusing on rider recruitment tactics. That makes sense as Uber started out as a mobility business – if it moves, they wanted to move it – so for them, the riders were key.
For Deliveroo, it was about good quality food so having the right selection on the platform was key for them, so they were starting at different points. With Uber entering, it started to converge quickly. The riders became commoditized because it became a job and people were on multiple platforms, so the battle became about selection.
In the long run, do you think selection is commoditized?
Yes. The Restaurant Group, the UK’s biggest restaurant chain, now have 12 or more virtual brands that they’ve created and you’re also getting into a lot of mirrored brands. To maintain lower platform fees, you’ve got a lot of businesses that will launch a virtual brand on one platform, as an exclusive brand, they launch exactly the same brand on a different platform under a different name and it’s exclusive on the new platform. The platforms themselves know this is happening but I don’t think the customers are necessarily aware of it. They look like they’re exclusive content on the platforms and I don’t think that’s a sustainable strategy, because anyone trying to build a quality brand over time and have customer loyalty, you’re not going to dilute your brand by having the same product sold on a different site with a completely different label, so I don’t think that’s sustainable. In the short term I think it’s interesting for brands. Basically it’s a play to say the fees are too expensive and this is a way around them.
Let’s talk about the marketplace model versus logistics. You had Just Eat with big margins, serving those UK restaurants that deliver themselves and then Uber and Deliveroo come in with a logistics network and offer the delivery component as well as demand from their customers. What were the biggest challenges you saw in Deliveroo launching Marketplace+ or the logistics models going into marketplace and competing with Just Eat?
I think the starting point is, when you go into the pure marketplace, you’re taking a phone directory and shifting it to an online business, so the barriers of that are relatively low. They’re not nothing, but they’re relatively low. It’s more about customer acquisition, making it a seamless customer experience, orders and get the selection on. When you’re starting in the logistics business which is where Deliveroo are trying to pioneer this in the UK, you spend a lot of money building out the infrastructure. You spend a lot of money educating customers that are ordering from good quality restaurants. You spend a lot of money educating restaurants that this is not cannibalizing their sales, it’s incremental. You also spend a lot of money educating riders to say, this is a real job and then building out the technology. A lot more investment goes into getting that business off the ground because you are building infrastructure and if you look at when you have something like Uber Eats enter the space, all of a sudden you’ve got two people building the infrastructure, so the net costs start to become cheaper and you start to commoditize it.
The interesting part is when you’re starting out and you’re spending all this money, you can’t get your customers to pay for everything, so you share the costs which means you have higher delivery fees for the customers and higher platform fees for the restaurants. I think what hasn’t happened yet is that now that those things are getting built, you’ve seen the delivery fee still going up and the marketplace fees to the restaurants have stayed at the same level. If you look at tech, for example if you invest in a hardware business, the first time you buy new hardware, it’s really expensive because you’re paying for the R&D, but over time it gets commoditized and cheap. So compared to the food delivery space, it’s now getting to the point where the delivery part should be becoming really cheap and we’re not seeing that. We’re seeing consumer prices going up and an unchanged restaurant price. So I do think there’s going to be a reset at some point, in terms of that margin structure, because food delivery should be a low margin business.
Why is the price going up? Do the logistics players not have the density required to cover their economics?
I think that’s been a starting point. It is a density play, you want to increase density, that’s one part of it and so when you start looking at pricing to consumers based on distance travelled, that’s a great mechanism for trying to increase the amount of volume you do closer to the restaurant, which over time allows you to start shrinking the zone. I think it’s a smart thing to do by pricing that way, because there’s a cost of service over distance. Customers can self-select how to do that. The advantage of that also means, if you can drive more demand closer to the customer, over time you can actually get more volume per hour through your rider base, then you can negotiate on the cost per hour. I think it’s a good driving force to get to an end state where net net you end with a better service, at a lower price to the consumer and you’re driving more density into the market.
The other piece you touched on was where you look at a marketplace versus a delivered service, eBay versus Amazon for example, because I spent so much time there. When eBay started out it built this incredible marketplace but the barriers to entry are a lot lower than if you’ve got physical infrastructure, where you’ve invested in fulfilment centers and logistics and inventory.
The network effects of eBay is that it fuels one another, so how do you look at disrupting that?
If I look at how that plays out, if you look at how much of the journey you take control of. Let’s say if Deliveroo sets up a restaurant, it has 100% control over what restaurant it sets up and how it’s priced, its location and the number of outlets. It has complete control so the further down the supply chain you get, the more control you have over the customer experience. If I look at a pure marketplace where Just Eat started, very minimal control over customer experience: can’t control restaurant price and food delivery speed, so very low control over the platform. Deliveroo put itself in a slightly higher control position which is, you can’t control the price but you can control the customer experience in terms of delivery time, so it’s a slightly higher control business. You’re seeing that both Uber Eats and Deliveroo have got into their own restaurant spaces now which is the final piece of building their own concepts where they can start controlling selection and pricing.
Which is the equivalent of Amazon going into Basics or private label products and shipping it through their network to consumers.
Exactly. All retailers have done this with own label, as Amazon did with Basics, so it’s a common thing to do. If you want a good customer experience the ideal is to play in all three spaces. You can generate good cashflow at low risk through the marketplace and you can reinvest some of that cash back into building out a logistics network, because these are hyper-local logistics networks. This then allows you to support a national distribution for chains, allows you to open up to cities or locations that otherwise you might be overinvesting in. The last piece is then, once you’ve got good demand and infrastructure, you can then launch your own restaurant in that area. This strategy has been used multiple times, it’s not a new approach. It’s just being applied to a different space.
Walk us through this. Just Eat typically own rural UK, so all restaurants outside big cities. Just Eat’s got something like 95% of UK postcodes covered. For Deliveroo or Uber Eats, for them to go into the rural areas and disrupt Just Eat, the big piece here is density. Does the density and logistics work in those rural areas?
The inputs are interesting, because density and volume are two things. Population density is lower but also speed is generally faster because of less traffic. Travelling speed in London, it’s high density and high traffic volume which is why you use bikes and scooters. If you start going into other locations, you can increase the radius. What you’re really worried about is the speed of getting to customers, not distance, so really, it’s a time question: how long am I on the road for to keep the food fresh? You can increase the radius. I think that’s one side that you can do in a rural area, to have a bigger catchment area.
The other piece is affluence. Outside of London, generally the more affluent you are, very quickly the lower the density gets. It can be different in London; the closer you’re getting, you’re increasing affluence, unless you’re getting into the suburbs. So when you’ve got an affluent product which is selling people food that is priced at restaurant prices, that’s the real challenge. If you’re in an affluent area, you can go to a restaurant every single day but the distance from home is vast, so it’s hard to maintain that good quality service and get density. If you’re going into less affluent locations, the reality is that people can’t afford to go to the restaurant every single day.