What do we want? Split the bill! When do we want it? We’re divided
Split, split, split, split… boom 💥
Let’s set the scene.
It’s Saturday night and you’re staying in with your besties – let’s call them Ross and Rachel – to binge watch Game of Thrones or to see if the new Great British Bake Off still weighs up (#sorrynotsorry).
You suggest ordering from Deliveroo and everyone’s up for it. You open the app, order swiftly from your favourite burger joint and then sit back and relax. By now you’re probably thinking: ‘How have I ended up paying for all this?’
Ross says he’ll grab some cash out for you later, but you’ve heard that one before. And Rachel says she’ll get the next one, but you’re sure she said that last time 🤔 – there must be a better way!
At Deliveroo, ‘split the bill’ is probably one of our most requested features, and it’s been on our roadmap since I joined the company over two years ago. In that time, we’ve done email surveys, phone interviews and multivariant tests to try and definitively prove its worth.
We also ran a really cool user research session by inviting a group of friends to our research lab (set up to look like a living room) and asking them to relax, watch Netflix and play video games while placing an order. This let us observe their group ordering habits and see how they split the cost between themselves.
We also ran an experiment earlier in the year to try and gauge how many people would use a bill-splitting feature. We added a ‘share the cost’ button to checkout, which let customers select how many people they wanted to split with. This was actually more of a survey than a functioning feature, as we wanted to discover how many people would use this and how frequently.
After undertaking all of these different research methods and experiments, we still have a relatively murky idea of how impactful a split the bill function would actually be. Even though people request it, the data seems to suggest that a lower number of people would use it, albeit with more frequency.
If you build it, they will come
So why not just build it and see how it goes?
Well for one, it’s a very big piece of work, both from a user experience standpoint and a technical one. There are many challenges and considerations, such as the time-sensitive nature of splitting the cost of something that only takes 20 minutes or so to arrive.
Do users want to split the cost evenly, based on a percentage or per item? What’s the best way of taking payment from multiple cards, including card fees? How do we handle both current customers and signing up new ones on the same order?
These are just some of the things we’d need to resolve. And taking all of this into account, we’d also need to compare the value to our customers against how impactful it would be to the business.
We had the idea to use Monzo.me for a split the bill function when they launched it in December last year. But at the time we thought it would require us to integrate with their API, so it was a non-starter.
Our team started talking about it again a few weeks ago and we came up with the idea to build a very simple solution by allowing customers to share their Monzo.me link from within the Deliveroo app. We knew we’d be able to learn so much more by having a functioning split the bill feature. As it was a quite a small amount of work, we pitched it to our product manager and then got the green light to go and build it, which is exactly what we did.
The solution is quite basic but functional. Customers can save their Monzo.me username on first use, then select the number of people they want to split with, including themselves. We calculate the cost per person and then customers can share their Monzo link via the standard OS share sheet. The amount is pre-filled in the link along with the restaurant name, which creates a link like this:
(feel free to send me money 🤑)
Then all the heavy lifting is done by our friends over at Monzo.
For us this is a lightweight experiment that has given us and our customers a functioning split the bill feature. Our next steps will be to observe it’s use, frequency, and if it helps to acquire new users. We’ll analyse the test results and either iterate to improve the feature in its current form or, if successful, build our own bill-splitting payments system.