Scaling Uber EATS in London

Can you share some context to when you joined Uber, specifically to your role running Eats in London?

I joined Uber right at the start of 2016. Uber had been on an absolute tear in terms of its valuation, tripling its value every year type thing. We’d become pretty hot property. I used to be a banker and I was kind of tired of that banking thing and Uber seemed super sexy, so I got my application in, got my offer and then joined at the start of the year. I joined a group called the Rest of UK Launch & Expansion Function. In early Uber, the general manager runs a region and then the operations teams are under that general manager do the day-to-day. It was very localized. It wasn't run through central teams; it was run through the guys on the ground and I joined the function for launch and growing the business.

I joined as operations manager for Merseyside which was Liverpool and the surrounding area, working with the GM for Merseyside and the Yorkshire region and our job was to make it work. That meant answering customer complaints, talking to drivers, doing analysis of competitors, doing pricing reviews, doing system config; anything that needed to happen came through us. In that nine months I was in the role, we tripled the business and it was the amazing, early halcyon days of Uber where everything was super exciting and everything was very unformed, you could try loads new stuff and every few months they’d invent a new search or dispatch algorithm and you’d talk to engineers and you’d roll it out in your area. Things were changing all the time.

Then quite sensibly, they realized that the business needed to consolidate, so what was happening is all of the best ideas were being invented in silos and never really shared and deployed more globally. If I’m honest, there was competition between us guys on the ground to be the guy that invented the cool thing and shared it and less excitement to take someone else's great idea and implement it in our region.

They did a horizontal to vertical pivot and they took us from owning just a region to owning one topic for the whole country. I went from Merseyside ops manager to looking after UberPool for London, so rather than looking after a region I just focused very narrowly on that product. Unlike the core rides business which was pumping all the time and growing really aggressively, Pool had a really difficult life in the UK, mainly because it was competing with UberX which is a phenomenal product. We came to market UberPool and said hey, it’s UberX but a bit cheaper and you share your vehicle. The British aren't super gregarious by nature and it was a bit of a challenge getting them to share. There's also weather and darkness challenges, so if you tell people to wait for a vehicle to pick them up, it's a lot easier in a climate where it's not raining and where the street architecture is more simplistic.

It was a real tough challenge. I worked there for a year and I saw a lot of improvement in the product; I worked really closely with the engineering team in the US to get new features rolled out in the UK. We really improved efficiency; we didn't really grow the product very much but we definitely made it work better.

Then that the time that Uber Eats was getting going. We'd had those early trials in the US as part of Uber Everything and then it was coming to the UK as a food-based product only. It had been launched but it was still a very small region of the UK that was being covered, London and maybe Manchester. I went and had a chat with that team and said, are there any roles going? It was a hilarious conversation because there’d been a real push for the Eats team not to be poaching people from the rides function. They didn't want everybody to come piling across and try and take a new role, so there was this kind of unwritten rule that you could only talk to someone who’d approached you first.

I had a conversation with an old colleague from rides. He got me in front of the GM that afternoon and we had a conversation about what roles were available and then I had to reapply. It was exactly the same as applying to Uber from the outside. I went through the whole six or seven interviews, case study and everything, apart from the numeracy test, I did from scratch. Then I joined this territories team, a really small function, eight of us and our job was to be on the ground: know the city, get to know the restaurants, get a feel for the local market dynamics. What we were to compliment was these larger central functions: restaurant function, driver function, marketing function in which we were trying to run the business from a scalable, do everything once for the whole country perspective.

What we learned, quite interestingly, is if you go too far away from that original Uber V1 model of people on the ground, general manager on the ground and you go to full centralization, you lose a lot of what made that strong. You gain a lot of efficiency in sharing of best practice and quick deployment but you lose a lot of local focus, urgency and ownership.

So we had this really interesting challenge, as the territories team, to move from a small component to almost co-owners of the business with the central teams. Over the 18 months where I was responsible for London we went from just me looking after London to hiring up a team of 20 people which I became head of, P&L owner, responsible for the actual performance of the business and partnering with the central teams to deploy either best practice or depths of experience to help solve our problems. Also being brought into everything from regulatory problems through to working with top accounts, like McDonald’s, through to looking at local marketplace optimization, pricing plans. Everything that might need a local lens was owned within the team.

That long journey through Uber was from local to central and then to team in Uber Eats that sort of scaled up to try and find that balance between being on the ground and being able to do best practice at scale.

You walk into Uber Eats at the end of 2017. What does the market landscape look like in London?

We’d been launched as long as anywhere else, but we'd come in probably against the backdrop of Deliveroo and Just Eat being really locked in a competition already, so we came in as a third party. Deliveroo had launched in the center of London. Their proposition was very much premium: better restaurants, better delivery times, nicer app. Just Eat was the traditional model of the restaurant that does its own delivery, but whatever they lost in terms of that premium, they made up for in terms of incredible selection, best price in the market and still a pretty good app and online experience.

So the two of them were kind of locked in this competition and Uber Eats came in and just went absolutely ballistic and started offering incredible guarantees. Our launch guarantees were like, if your meal is not there within 30 minutes you get a whole free meal again. We very quickly realized that wasn't sustainable. We had the 2006 mindset of throw money at the problem and we're going to solve it and we had to catch up very quickly with that 2018 mindset of the market actually wants to see that we’re growing in sophistication, rather than just chucking money at problems. We had to climb down really quickly from explosively large spend in a small area, to the much harder graft of we’ve got to start competing with Just Eat in terms of prices and selection and competing with Deliveroo in terms of quality of offering. That was against the backdrop of having to roll out across the whole of the UK, as well as just keep London ticking over.

Just Eat was mainly offering the marketplace or were they doing logistics as well in 2017, 2018?

At that point, if Just Eat were experimenting with that courier-led model, I'm not aware of it. I wouldn't want to give you the definitive answer but we certainly saw it as a clash of models back in the day. It was very much the more expensive, much quicker, much more controlled courier model versus the more freewheeling aggregator model. One had huge advantages for scalability, one had huge advantages for quality and it was before, I think, anybody had really recognized that both models had real strengths and that you were probably going to have to do different ones in different places. It was very much a world of each thinking their model’s best.

What was your strategy going in as Uber Eats in London?

The things that we wanted to bring to the table that were our unique advantages were on the courier side, we inherited this phenomenal technology from the core rides business. So we had the ability to give a driver app, dispatch to restaurants, have a great real time database, have all the tools to deal with things like bridges. Loads and loads of platforms get confused about geographical obstacles because you try dispatch a courier to a restaurant over a river and you can't as there's no bridge there. That sounds dumb but it's going to spoil half of London if you think about the massive river running through it. We had those tools and we had the experience, the analytical ability of all the people that were now in the team. That meant that, very quickly, we got to a super reliable liquid marketplace. That was one thing.

The second thing I think we brought to the table is global scale. We were able to come into the market and have conversations with the likes of McDonald’s, not as a little upstart but as a, ‘We are your global best partner. We're sophisticated. We operate in more markets than anybody else. If you partner with us in the US, you're also partnering with us in Australia and us in Europe and in Sub-Saharan Africa or wherever you want to be.’ I think that gave us a huge advantage versus the likes of Deliveroo and Just Eat because they couldn't have those kinds of conversations with Starbucks, KFC and McDonald’s. That isn't to say they weren’t a great local partner and they didn't speak to them locally, but our ability to go to a very high level at those organizations, internationally, was kind of unrivaled. That's what we brought to market day one: amazing partnerships and close to best in class technology.

In terms of scale, was it more important on the driver’s side that you had liquid drivers and riders that could switch from rides to Eats or was it harder to get the demand?