How to run a Minimum Viable Event

“I’ve put a date in everyone’s calendar. That means we have to organise something.”

That’s how preparations started a few weeks ago, for when the Deliveroo design team ran our first (Minimum Viable) Event! Soon enough, we were at ustwo’s London studio with a sold-out Eventbrite list, great speakers, and a table of pizza straight from Deliveroo’s finest.

Several people everywhere.

Before we got started, James gave a quick overview of how design works at Deliveroo. It’s been a huge year for the company, and the design team’s scaled to match, from two people to 17.

To borrow a phrase: we touch everything. Broadly speaking, we cover product and brand design, and that makes us responsible for Deliveroo’s apps, website, internal tools, photography, driver gear, bus wraps, flyers, newsletters and chocolate bars (yep). Everything — both print and digital — shapes our users’ experience, and as the company grows, we want to make sure we’re moving quickly too, while solving the right problems and validating the decisions we make. So we thought we’d invite along some people who help us make better design decisions, to explain what they do.

Step forward, Sofia from NomNom, Murat from Marvel and our very own JP. Each of them spent some time talking about what they‘re working on, before we brought everyone together for a fireside chat and a group prototype-a-thon. Here are four big things we learnt.


1. Know your users

We all know we shouldn’t design in a vacuum. NomNom is really good at helping us use customer insights to meaningfully improve our products, and it was great to hear more about it from Sofia. She noticed that people who make an impact on the business are often removed from the users themselves and so created NomNom, a platform to consolidate customer feedback and get to the heart of what it means. Deliveroo now uses it to learn from our customer service and Twitter replies.

Later in the evening, we asked our speakers how users of their products have continued to surprise them, despite the insights they have available. Murat talked about Marvel’s growth: as the product becomes more widely known, it’s becoming aimed not just at designers, but at people who want to make something. It’s a transition that‘s continually expanded the range of things that people want to build with it.

As for how Deliveroo’s users have surprised JP? ‘They order a lot of food!’

One of Murat’s previous client requests. We’ve all been there.

2. It’s often the side projects

Next up, Murat showed us several of the side projects that he began a few years ago while working at agencies. One of those was the first version of Marvel, for which investment and growth quickly followed. It’s now a platform we use at Deliveroo as a standard part of our product design process.

During our fireside chat, the topic of hackathons and side projects came up again (with Sofia describing NomNom as ‘like a ten-month hackathon!’), and it was fairly unanimously agreed that they’re well worth the time. As Murat pointed out: if you’re hiring, a banner ad to that effect will only be visible for a certain amount of time, but side projects can continue to work for you to attract talent.

(On a related note: within the Deliveroo product team, few can claim to have mastered the art of the side project better than JP. A real set of traffic lights that hang on the office wall and show the status of our codebase’s most recent build? Right this way.)

A quick fireside chat with JP, Murat and Sofia.

3. Practical beats fun, but you need both

Even as a tool for building quick prototypes, the Marvel team are well-known for paying attention to the delightful details. Take a look at the animation on the 404 page. If you become a paid subscriber, look out for the confetti. But while these are important, it’s Marvel’s sheer practical utility that deservedly gets it the most attention.

It’s what we aim for with Deliveroo’s design. Generally, you can rely on ugly but functional products to win out against less useful ones that happen to be prettier. Balancing utility with delight can be tricky, but getting them both right is something the Deliveroo design team are intensely focused on.

4. Keep making sure you built the right thing

Last but not least: JP, one of our brilliant consumer team engineers. Deliveroo’s at the size now where data can really help influence our design decisions, so we can use it to answer the question, ‘did we build the right thing?’

If you’re a Deliveroo website user, you may have noticed our list of nearby restaurants being presented as images recently. In which case: congratulations! You’ve been part of a (successful) A/B test to see whether it would increase conversion for those 50% of our users.

Combined with our process of continuously deploying new code to our servers, we can understand faster than ever whether we have, in fact, built the right thing. And if the answer turns out to be ‘no’, we can revert it and, increasingly quickly, design something that is.

All that said, to what extent do our speakers allow their own teams’ views to influence their roadmaps? Marvel removed features that didn’t fit into their product vision, since keeping them around could make newer ones more difficult to build — and they deliberately don’t pay too much attention to competitors. The data’s hugely important, but sometimes the people who are building your product are the ones who know it best.


“They’re waterproof goggles. So… goggles.”

But how do we work out what issues our product designs should address in the first place? It depends, but defining job stories (“as an X doing Y, I want to be able to Z.”) is often a good start. So, last thing of the evening: get into teams, be allocated a random job story, and solve it in 15 minutes.

“As a detective grinding coffee beans, I want to be warned of danger.”

You had to be there, but if you’re a musically-inclined farmer, a magazine-reading rubber fetishist, or Matt Damon who’s finally bored of being stuck on Mars… did we all think up a solution for you.

Channel-swimming Martians who need to check their texts: rejoice!

Then, the pub.

And that was #mve, our evening of hearing from people who are building things that help us be better designers. Huge thanks to ustwo, Sofia, Murat, JP, and everyone who came along. And if you want to hear more about what we’re up to, come and say hi on Twitter.

We’ll be back with another Minimum Viable Event, so see you there! Now we just need to put that date in the calendar…