Uber Money with Ebi Atawodi (Big Bets 9)
This is a newsletter about big bets, explored through conversations with the product leaders who worked on them. For this one, I chatted to the awesome Ebi Atawodi about two big bets she led at Uber: how they built a wallet in Uber, and how they started building a bank for the gig economy. One of Ebi’s bets came off and the other didn’t, but there’s as much to learn from each.
Money problems
Ebi describes herself as an intuition-led product manager. But, she points out, intuition comes from experience, it’s a muscle you build. Two bets she led in her time at Uber provide a great example of this.
They both solved money problems, and both came from the intuition built out of a deep understanding of Uber’s customers. ‘Every time we’d go and do UX research studies, we’d always include a few foundational questions to try to go deep on the context of the driver,’ she explains. Across these studies, in countries from France and the UK to Brazil and Indonesia, a few core themes started to emerge. For example, a large number of drivers both earn and spend with Uber. They said they were worried about not having a safety net - in the US 40% had less than $400 available - what would happen if they got a flat tire and couldn’t drive to earn the money they needed to fix it? And they had a lot of unfulfilled aspirations - two common ones were taking their family on holiday, or getting a mortgage. As a next step in life, many wanted to return to education, or start their own business.
These insights matched what Ebi herself had seen as the General Manager for West Africa from 2014-16. When they launched, many prospective drivers didn’t even have a bank account (globally that number is 35%) or experience using a smartphone - so she’d partnered with a bank to help them set up their first account, and partnered with Google to run training on using Google Maps. These drivers would often move from renting a car, to owning a car, and then maybe even owning another one and renting that out. It wasn’t until a few years later, in 2017, when Ebi transitioned into the product team, that she had a chance to bring all these insights together. ‘You don’t know when you’re going to stumble onto the problem,’ Ebi says, ‘so keep digging for it, to build that intuition.’
Ebi took over a product team called Money Hub, and resolved that at any given time everyone in the team should know what the top 10 user problems were for their product area. They created a shared doc called ‘Money problems’ (playing on ‘More money, more problems’) which captured the insights they were getting from their user research, support tickets, and data insights across all teams.
Problems > ideas
When Ebi decided it was time to turn some of these problems into ideas, she didn’t yet have any funding to do so. She did, however, have a Product Marketing Manager (PMM) called Adam Weigand who was keen to work on this, so they decided to kick it off scrappily. They worked with a UX researcher to help design an interview guide to validate the problems. Then Adam booked a few flights to Brazil, Mexico and the UK, and just stopped people in the streets to ask them the questions. Adam created an internal Google+ page and, like an internal influencer, would write posts any time he came across something relevant, e.g. ‘I just met this driver and he said…’. They were keen to make others as excited about these problems as they were. Ebi says: ‘It always takes time to synthesize research after the fact, so it’s good to create breadcrumbs for people to follow, and let them join in.’
Armed with validation and some fresh insights, they then decided to figure out what the marquee use case would be for each problem. Ebi believes that you should be able to express any marquee use case as though it were a billboard ad, describing the core problem and the core product in a sentence each.
At the time the team had no-meeting Wednesdays, so Ebi and Adam reserved 3 in a row for working on this project. Every Wednesday morning, they found an empty room and started storyboarding with post-its, not stopping until the end of the day. They picked an archetype persona from a relevant country, and then they mapped out a user journey describing a solution to each of their problems. They generated about 14 or 15 of these stories, including a journey for sending money overseas more cheaply and easily with the Uber app, and a journey for getting a temporary loan from Uber to help pay for urgent car repairs.
After those three weeks were up, they then transcribed these stories into Google Slides, and kept adding to them over time with ‘ephemera’, in Ebi’s words. These were little things - a screenshot or an email - that brought an element of the story to life, e.g. a brand from which they could secure discounts or a UI that inspired part of the user journey.
They also found another team whose work aligned with theirs: the Loyalty team. The Loyalty team were trying to build a driver loyalty programme and were inspired by the Money team’s storyboard approach. They created lots of user journeys for rewards, such as ‘I’m able to get a free tire change at [x] store’.
6 months after this project started, Uber hired a Head of Money, Peter Hazlehurst. He arrived, and started mapping out his entire vision and coming up with revenue models. He saw a world where Uber built the bank for the gig economy, touching on the exact same Money Problems. This is not that unusual, Ebi points out - especially in a large company there are often other people in the business who will arrive at the same or similar conclusions as you. Delighted to have an ally, Ebi reached out to him and offered to show him what they’d been working on. Peter looked at all of the user stories and was impressed - he agreed they were on to something, and told them to ‘go do their thing’.
Prototype > launch
From there, the bet started to accelerate. They pulled in a designer who very quickly (‘one day in Figma’) made a flow for each of the story boards. Each storyboard reflected one of 14 marquee products. Ebi picked out the 2 or 3 that best told the story, Adam created a single minded message for each product, and they brought those to high fidelity, putting them into a vision doc that they aspirationally called Money 2020 - their plan was to evangelise this story, and then launch Uber Money at Money20/20 in Vegas (the biggest fintech conference). They now had a date to work towards.
At the time there was no surface in the app specifically dedicated to managing your money. In fact, drivers’ trip activity was tightly coupled with their money movement, which made it hard to see what they’d earned in total. The analogy Ebi came up with was that money movements should operate like your bank account, but your activities are more like strava - you want to know how you did, what your pace was etc. Or, to put it another way, your paycheck should be presented separately from your performance review.
The logical answer to this was to build a wallet within the Uber driver app which bifurcated earnings and money. Ebi liked the name Wallet, because she’s a big believer in products that have a tight association to real life objects that people understand - it can clear up a lot of confusion around new concepts.
In order to build the Wallet MVP, they put together a small team ad hoc. Ebi was the product manager, and she got her lead cross-functional partners (design, data, engineering managers) to each contribute a few folks until they had a team of 6 (design, UXR, content strategy, data, two engineers). Their plan was to see what happened, and then use that to help them decide whether to fund it properly. ‘You need to have implicit trust with your cross-functional partners in the early stages of a new product,’ Ebi says, explaining how she was able to muster the team.
The driver (‘earner’) wallet prototype shipped in 2019. It had just a few use cases, primarily focused on better discovery of things you needed to do with your money, such as add a bank account, finish your KYC, and update expired cards. Even this very limited version showed promise - the wallet led to a 15% increase in the number of new drivers who added their bank account. This was significant because one of the biggest reasons for driver churn in the first week is not having added a bank account - without a bank account you can’t get paid.
Uber had launched a debit card in 2017 but the Money team and Loyalty teams relaunched it within the Uber Wallet. They also tied it onto a $100 fee free overdraft (known as a ‘back-up balance’), to help address the insecurity they’d heard in the original research from drivers who were worried their car would break down and they wouldn’t be able to earn money to fix it. 56% of eligible drivers were taking advantage of this, and 60% of them were using it six or more times a month.
At this point the team also started working on a rider (‘spender’) wallet. There was already a product called Uber Cash, which was a stored value balance, but it didn’t have any history of spending. 40% of Uber Cash support tickets at the time were actually based on lack of awareness of the balance. Much like the drivers, riders were confused about how money worked or rather ‘moved’ in the Uber app. So Ebi and the team started very simply, they bootstrapped a team of two engineers and on the top right corner of your homescreen they added a pill showing your Uber Cash balance, which took you to your spending history (the forerunner of the wallet). It instantly drove a 10% increase in usage of Uber Cash.
Now there was a live prototype of both the driver and the rider wallet, and each had proven that they had promise. So the team decided this was the point to go and tell the story of what an actual wallet could look like. Their original designer started mocking up a fully fledged wallet with all the marquee use cases. Armed with this vision, Ebi and Peter secured funding for the project and started designing their org. They even went on the company all hands and presented the vision of ‘one Uber, one customer, one Wallet’, despite the fact that most of these products didn’t yet exist, ‘but it helped with internal recruitment,’ Ebi laughs. They categorised the 14 products into 4 groups, and Peter staffed a team on each. The big bet on money was now a real thing, and you can find the Uber Wallet in your Uber app today.
Project Cheddar
But that’s not where the story ends. One branch of the money vision had been the wallet, which was a realistic feature that could ship and delivered instant user value. But the broader vision was that money would become a competitive advantage for Uber. And for that to happen, Ebi believed they needed to go further and create a banking app for Uber Money - in fact, a bank for gig economy workers. This initiative became codenamed Project Cheddar.
Once again Ebi went to the team leads and asked them to donate a person to this new project. Every manager found one, and because they were inspired by the vision, they each sent their best person. In total they amassed a squad of 11 people from different teams, who were tasked with figuring out what they should build.
Within Uber, people were sceptical that the world needed another Uber app, but research very quickly validated the idea that it would be useful to have a standalone financial app. Drivers said they always had their phone on while they were waiting for a trip, often hanging out with other drivers, but they wouldn’t want to leave their bank account unlocked in reach of passers by. In addition, the team deliberately did a design projecting what the main rider and driver uber apps would look like if they carried on cramming financial services into them - it was a frankenstein. These two artefacts of the research and monstrous design were enough to convince people of the need for a new app.
They created a high fidelity prototype of how the app might work, and had drivers go through it. The model they came up with was ‘view and act’. ‘The richer you are, the more stable your finances are and the less you ‘view’, i.e. check your balance,’ explains Ebi. ‘But someone with low income goes into a store, checks their balance - view, and asks themselves if they have money for this purchase - act.’ Traditional bank accounts are not designed for the gig economy, or people who earn in a dynamic way. So there was a clear need to build a bank for the gig economy, based on this idea of regular viewing, and giving users the right information, such as what bills are coming up.
The prototype was a hit, so they started building it for real. They planned to start by shipping a beta it to 100 drivers. But then covid happened. Uber downsized, cutting 25% of their total headcount, to deal with the massive drop in demand for rides. Suddenly they barely had enough people to cover fundamentals, and all bets projects were stopped, including Project Cheddar.
It’s clear that Ebi is proud of what the team achieved, but disappointed about the things they didn’t have a chance to try. ‘Even now people from those days still call me and say: Project Cheddar was the most exciting thing I’ve ever worked on,’ she tells me. Some bets are a bad idea, or have bad execution; others just suffer from bad luck.
My top takeaways:
It sometimes seems like great ideas come in a flash of inspiration. But if you really want to innovate, invest in building your intuition about users through user research, data analysis, hands-on experience or whatever else you have available.
Even in a big / well-resourced company, it pays to be scrappy with a big bet and create evidence and internal momentum before you pitch for a team on your new idea.
Use storyboards and product designs to build support for your ideas - these could be ‘good’ designs that communicate your vision or ‘bad’ designs of what might happen if you continue down the current path without taking this bet.
Sometimes great bets don’t work out for reasons beyond your control. Keep your head up and always let the user be your guide.