An Interview with Sophie Woods, UX Researcher at Deliveroo

How did you get into UX Research?

Like a lot of graduates, I didn’t have a strong sense of my destiny after leaving uni. I’d studied psychology and I’ve always been interested in unpicking human behaviour and motivations. I also knew I wanted to work in a role that relied heavily on communication – real face-to-face communication. I love listening to people and talking to people, far more than I enjoy typing on a keyboard!

When I applied to work at Deliveroo for a Customer Insights role, I’d never heard of the company… and nor had anyone else. They seemed to like my enthusiasm and weren’t too deterred by my lack of experience. The next thing I knew it was my first day on the job.

The role was sort of like an informal user research role within the operations team. I spent my days talking to riders about the pros and cons of the Deliveroo kit or organising focus groups with customers who’d had a bad experience. I was the only person doing this so I was figuring it out as I went along. It was scrappy and exciting!

About 3 months after I joined, Deliveroo hired Charlotte to build and lead the user research team. We ended up chatting over coffee about the things I was working on. I think she could see that I was eager to understand our users better but lacked any formal training or experience in research.

Amazingly, Charlotte took a gamble on me and offered me a role in her team as the first junior researcher. I was thrilled. We were the only researchers at Deliveroo for the following 8 months. I consider these 8 months to be the crash course in user research that set me up for the next 3 years at Deliveroo. It was during there 8 months I realised I’d found what I was looking for!

What does your role entail as UX Researcher and what are your key goals?

A ‘day in the life’ doesn’t really work for a user researcher at Deliveroo. For instance, today I’ve been phoning restaurant owners to ask about the hardware challenges they face, but a few months ago I was following a Deliveroo rider around the busy highways of Taiwan on a moped! (with consent, of course!).

Those are the more glamorous moments, but I spend most of my time with my product team figuring out what we don’t know, as well as feeding what I do know about our users into the work the team does.

Ultimately, I believe a researcher's goal should be to ensure the people in your company, at whatever level, are able to make decisions with a strong sense of the user's needs, and of the potential user impacts.

What's the structure of your team like?

What's your process like for each new project?

Research questions can come from anywhere – from unexpected trends in data, our product team's quarterly goals or us simply realising we know nothing about something.

Depending of the scope of a project, the process looks totally different. We can usually turn around usability testing pretty quickly.

Our process here involves working closely with the product and content designer, and often PM. These evaluative projects tend to be quite informal with light-touch project plans and analysis.

When it comes to foundational work – where we’re answering relatively big and complex questions – the process takes a different form. We’ll usually host a kick-off meeting with stakeholders to explain why we’re tackling this problem. In this meeting, we’ll understand their hopes and fears of the project. We’ll then crowdsource research questions from the group. It’s best to know everything they want to learn upfront so we can accurately set their expectations.

Next, we’ll look into the data on the topic to identify which markets to target. The research team is based in the UK but our remit is global, so we consider all of our markets when we kick off a project.

Alongside this we’ll probably be doing some desk research to collate what we already know about the topic. This might mean flicking through old transcripts, looking at experiment results, or just snooping around on Google for anything relevant.

Once we have our research questions, we know the markets to focus on, and we’ve gathered all the context we can, we can start to think about the details of the study designs. From here on, every project looks different!

When there are multiple stakeholders on the project we usually establish a pattern of communication upfront such as weekly check-ins, to ensure everyone knows what to expect and that everyone’s kept in the loop and engaged with the progress.

How do you work with the design team?

I feel like I’m in a love triangle with content design and product design! We work extremely closely with both disciplines. Our wider team is called CRD which stands for Content, Research, Design.

Due to our embedded model, it's easy for the 3 disciplines to kind of move together in unison. For example, we might want to concept test a few new ideas through prototypes. To kick off a project like this, all 3 of us might sit in a room to discuss the concepts and the style of testing that would work best.

Then the product and content designer will develop the designs while I am planning the research e.g. recruitment and facility hire. To an extent, the research session itself is co-designed by all three of us – lead by the researcher, but with content and product design input to the discussion guide. Then on the day, we’re all present. Content and design are usually chief note-takers – you’ll find them furiously post-it-ing.

For the most part, the analysis is taken on by the researcher. But, for me good collaboration is when you don’t really need to spend time putting together a ‘deliverable’ from your research, because your CRD partners have been on the journey with you and have just as detailed an understanding of the insights as the researcher (of course, we document our findings anyway!). I try to think of the researcher as the facilitator of the team having knowledge on the user, not the owner of the knowledge of the user.

How do you communicate insights to the rest of the business?

At the risk of sounding obvious, bring the people to the research and if that can’t work, bring the research to the people!

"If you predict you’re going to have trouble convincing a certain set of people to action the findings, there’s justification to compromise on the study design to some extent to enhance engagement, while being careful your research still maintains integrity."

For example, imagine the optimal method for your project is in-home interviews because you want to understand how consumers store and cook food at home. However, getting engagement from stakeholders is tough because space to observe is limited, travel times are lengthy and timings are specific. In this case, I’d try to find a way to make the research more accessible. For instance, asking participants to come into a lab near HQ and to replace the in-context element, run a light-touch diary study the week prior, where your participants take videos of their kitchen and photos of the food they’ve cooked. These will act as talking points in the lab session.

This way stakeholders can easily drop in to observe research, and half the legwork of ‘communication’ is done before you even start the analysis. After all, a rigorous and technically perfect project isn’t all that much use if the stakeholders who need to action it aren’t engaged.

“When it comes to the research output, how it’s communicated is so important. Design your output with the user in mind. Unfortunately, a one size-fits-all approach doesn’t work. This can be time-consuming but it can also be the difference between findings that land and findings that don’t.”

For example, do the executives need an in-depth analysis? Probably not, they want the take-homes and the action points. Whereas your product team, they probably need the detail to understand exactly why, and how to action it.

“I strongly believe you need to sculpt your communication style and your outputs for the people consuming it.”

What does Research ops mean to Deliveroo?

Our ResearchOps team is still relatively new and is made up of two wonderful people. Their aim is to give us researchers the support and infrastructure we need to scale our impact effectively as the business grows – essentially they help to operationalise what we do.

This support comes in various forms such as handling our participant recruitment, research planning and logistics, helping us to find tools to analyse and store our research more effectively, and organising guest speakers to help educate/upskill. All sorts really!

Research ops have made my life so much easier, as a team, we have a lot to thank them for. Also, thank god they came when they did, I’m not sure we’d have survived GDPR without them!

If you’re interested to know more, here’s a blog post written by Saskia, our research programme manager about her first year in the role!

What has been the biggest challenge you've overcome working in UX research?

Any advice on best practice when testing with users?

The user is the expert and you are the novice. You should show them you’re there to learn from them, and that's it. This means avoiding any technical language, adopting their language, and asking the most basic of questions to show them you’ve made no assumptions. If the participant feels like they are more knowledgeable than you about the problems, I believe they’ll give you more.

“The user is the expert and you are the novice.”

Additionally, to build real rapport and ultimately unlock the nitty-gritty, you need to show them you’re a human too. Don't be afraid to laugh along, make a joke (not out of them), mention your weaknesses, for example “I’ve also ordered a burger before 11am on a Sunday.”

I used to be a bit more sterile in my approach, to avoid any risk of projecting my personality onto the environment. Then I realised you can still have a personality without biasing a session. So go with the flow and don’t be rigid with your guide.

“I think finding the art of a good research interview isn’t too different from finding the art of good conversation. Listen, express interest, be open and be ready to learn. These people honestly know the product better than you… at least when it comes to being a user.”

What's the best advice you've received?

My manager told me a researcher should be researching 30% of the time and communicating research 70% of the time. Before I heard this, I’d often feel frustrated by the number of things in my diary that got in the way of me just ‘researching’. I wanted to be out in the field all the time because that's my job.

But hearing this advice made me realise that if you spend five days a week interviewing, that doesn’t make you a good researcher. You might well be sitting on heaps and heaps of gold but unless it's being pumped back through your team, it's useless.

“You might well be sitting on heaps and heaps of gold but unless it's being pumped back through your team, it's useless.”

So when it comes to socialising your findings, don’t be afraid to overshare and be repetitive. One powerful insight shared 10 times is more impactful than 10 interesting insights shared once. I’m exaggerating, of course, but I hope you see my point.

“Being a pro-active, vocal member of your product team is being a researcher. Incorporating research insight into every conversation your product team’s having is being a researcher.”

Appreciating this has really changed my perspective and my priorities. I’m not saying stop researching all together – you’ll still need some of that, 30% to be precise. But communication is the most important thing.

Whose work do you admire or follow?

I love Louis Theroux. Obviously, the topics he investigates are fascinating in themselves. But, as a researcher, it's his interview style that really gets my attention.

He has an amazing knack for unlocking the people who you’d expect to be the most private. I guess it’s a researchers dream to be able to build rapport with a person whose world is a million miles away from your own. And he achieves that every time.

Through hours of what I like to call ‘studying’ (aka Netflix), I’ve noticed he comes across as naive, from his back-to-basics questions to his socially awkward, ‘I’m-too-tall-for-this-space’ body language. I think it’s this naivety that makes the viewers really root for him and the interviewees feel comfortable unravelling in front of him.

He shows true interest in the subject’s lifestyle, without a flash of disapproval. Which can't be easy when you consider the sort of environments he exposes himself to. And it really works, people reveal everything.
Regardless of whether his gawkiness is legitimate or deliberate, it's genius.

What do you see for the future of research?