How contextual research helped us redesign the replenishing process in our Food stores

Every day, in every Co-op Food store, a colleague does a ‘gap scan’. They walk around the store, they spot gaps on the shelves, and they scan the shelf label with a hand-held terminal. This generates a ‘gap report’ which tells the colleague which products need replenishing. It also flags other tasks, such as which items need taking off the shelves because they should no longer be sold.

This is an essential stock management process in our stores. It ensures:

  • stock we’re low on is ordered automatically
  • customers can get the products they need
  • our stock data is accurate

However, the process is complicated. There’s an 18-page user manual explaining how to do it and on average, gap reports are 25 pages long. 

Making the essential less arduous

In the Operational Innovation Store team, we aim to simplify laborious processes in stores. Product owner and former store manager Ross Milner began thinking about how we might tackle ‘gap’, as store colleagues call it. 

He started by asking some questions:

  • How might we design a process so intuitive our store colleagues don’t need a manual? 
  • How might we help colleagues complete all the priority actions from the report immediately? 
  • How might we save 25 pieces of paper per store, per day – in other words, 22 million sheets per year? 

Learning from users

I’m a user researcher and this is the point where I joined the project. My first research objective was to discover how store colleagues go about the process at the moment, and what they find good and bad about it. To do this, I visited 5 stores. I interviewed the managers about their process – as it’s a task which usually falls to them due to its current complexity – but most importantly, I observed how they use the gap reports.

Adapting what they had to meet their needs

Being there in person in the back offices in stores gave me a far deeper insight than I would have got had I done phone interviews, or even just spoken to colleagues on the shop floor. 

Being there gave me access to reams of old gap reports stashed in the back office. It was invaluable to see how colleagues had adapted them to better meet their needs. Some of the things I saw included:

  • dividing the stack of pages into easily-managed sections
  • highlighting the information that requires action
  • ignoring all the non-actionable information on the report – some users didn’t even know what the information meant
  • changing printer settings to save paper
  • ticking off products as they complete the actions against them 

Photograph of one page of a gap report. Several numbers are highlighted. Not particularly easy to understand.

Seeing the physical artefact in its context revealed a lot of needs we might have otherwise missed, because colleagues are doing these things subconsciously and most likely wouldn’t have thought to mention them to us.

Learning from prototypes

Our contextual research has helped us identify several unmet needs. Delivery manager Lee Connolly built a basic prototype in Sketch and we mocked up a digitised gap reporting process. The design clearly separated and prioritised anything that needed store colleagues to take action. We arranged those tasks in a list so they could be ‘ticked off’ in the moment, on the shop floor.

Screenshot of an early prototype used for scanning labels on shelves

This was intended as a talking point in user interviews and the feedback was positive. The store managers were fascinated, asking when they’d be able to use it, and – unprompted – listing all the benefits we were hoping to achieve, and more.

Developing ‘Replen’: an alpha

We’d validated some assumptions and with increased confidence in the idea, we expanded our team to include a designer and developer so we could build an alpha version of the app. We call this app ‘Replen’ because its aim is to help colleagues replenish products when needed.

Interaction designer Charles Burdett began rapid prototyping and usability testing to fail fast, learn quickly and improve confidence in the interface. It was important to do this in the store alongside colleagues, on the devices they normally use. We wanted to make it feel as realistic as possible so users could imagine how it would work as a whole process and we could elicit a natural response from them. 

photograph of possible interface on a phone in front of co-op food store shelves

Profiling stores so we know where we’re starting from

Before we could give them the app, we needed to understand each trial store’s current situation, so that we’ll be able to understand how much of a difference Replen has made. We visited all the stores we’re including in our trial. Again, being physically there, in context, was vital. 

The following things have an effect on the current gap process and may also affect how useful Replen is for colleagues. We noted:

  • the store layout and the size of their warehouse
  • whether the store tends to print double-sided
  • where managers had created their own posters and guides to help colleagues follow the gap process
  • any workarounds the stores are doing to save time and effort

Screen Shot 2019-07-01 at 16.25.04

What’s next for Replen?

We’ve just launched the Replen alpha in our 12 trial stores.

The aim of an alpha is to learn. We’re excited to see whether it meets user needs, and validate some of the benefits we’ve been talking about. We’re also keen to see whether stores continue using any workarounds, and whether cognitive load is reduced.

We will, of course, be learning this by visiting the stores in person, observing our product being used in real life, and speaking to our users face to face. When redesigning a process, user research in context is everything. 

Rachel Hand
User researcher

Matching our research approach to the project

We’ve been drafting our user research principles recently and one idea to come out of the sessions was that:

User researchers shouldn’t fall back on a single research method just because we know it.

It got me thinking about my latest project at Co-op Digital and how interviewing people in a lab may have been easier for us but our insights wouldn’t have been anywhere near as valuable.

My point is: it’s easy to stick with a research method because it’s familiar or we feel confident using it, but different projects demand different approaches and user researchers must think carefully about choosing the most suitable one.

Matching the research approach to the project

I’ve been working on a ‘later life planning’ team, looking at how we can help people plan for the future.

We wanted to have conversations with people around:

  • what planning for the future means in practice
  • their hopes for the future
  • any plans they have in place

But talking about wills, funerals, loss and what might happen to us in the future can be scary and emotional – so much so that it’s a conversation lots of people avoid having. I quickly realised that our research approach needed to be carefully and sensitively planned.

We use research labs regularly and they can be brilliant. They help the whole team witness the research first-hand and the controlled environment allows the researcher to focus on the interview rather than on logistics.

But labs can be quite clinical.

The white walls and huge two-way mirror don’t foster a comfortable, relaxed environment. I wanted the people we spoke to to be comfortable. This is something that needed to be more on their terms.

Researching in context

Photograph of hands cupping a mug at a dining table. Glasses in shot as well as a plate of chocolate biscuits.Apart from putting them at ease, the decision to visit people in their homes came from the desire to understand a wider context. What environment are people in when they have these conversations and make these decisions?

And home visits were great. Talking to someone surrounded by photos of their grandchildren, or with their pets bouncing around, helped us to understand what’s really important to them. By allowing us to see a little bit of how they live, our research participants gave us insights that we might not have got if we’d spoken over the phone or they’d talked to us in a lab.

For example, we saw:

  1. People struggling to find certain documents, despite them telling us that everything was in one place. This indicates that they might not have their plans as organised as they’d made out. We’re less likely to have found this out over the phone or in a lab.
  2. People’s expressions and body language while they had candid conversations with loved ones. One valuable insight was seeing the sense of urgency on a wife’s face when she spoke about needing to replace her husband’s expired life assurance plan. Her expression gave us an idea of what an important and worrying issue this was for her. Their body language show us how much importance each of them placed on different parts of their existing plan.  

Seeing things first-hand was a good reminder that people’s lives are messy. Anything we design or build needs to consider this.

Challenges with timing and practicalities

It took a long time to plan and prepare for the home visits. The logistics of travelling, getting lost, finding somewhere to park and finding the right spot to put the GoPro so we could record the interview was sometimes tricky. And by the time we’d introduced ourselves, talked through consent, set up the GoPro, things felt quite rushed. Next time, I’ll allocate time for these things or chat over the phone before the interview to establish a relationship and cover the basics in advance.

It’s tricky to involve the whole team

I’ve found it’s easier to get the whole team involved when we’re speaking to people in the lab – it’s one place, one day. But we carried out home visits over 2 weeks making it more difficult to pin everyone down.

We discussed who we wanted to speak to and what we’d like to find out as a team beforehand. This then fed into the plan and discussion guide. Product manager Sophia Ridge and designer Matt Tyas were able to come to the various interviews but making sure the whole team heard the voice of the interviewee and got the same insight was difficult.

We all came together to watch the home visit videos and I asked everyone to take notes as they would in a lab setting. I’d hoped we’d sort the findings and uncover themes and insights together. But 2 videos in, people were pulled onto other work. Next time, I’ll take the team out of our working space and ask them to leave their laptops behind.

Research community, how do you do it?

We’d be interested to hear how and why you’ve chosen to step outside the lab for different projects, and whether you think you got more useful insight from it. Leave a comment below.

Vicki Riley
User researcher

Funeralcare team welcomes a new user researcher

Photo of Mark Branigan, user researcher on Funeralcare

I’m Mark and I recently joined Co-op Digital as a user researcher on Co-op Funeralcare. I’m part of a multi-disciplinary team which means I work alongside developers, designers and product managers as well as internal and external subject matter experts from the funeralcare industry.

At Co-op Digital we’re building a service so that our Co-op Funeralcare colleagues can meet the needs of their customers more efficiently. The less time spent doing paperwork, the more time they have for their customers. Their role includes welcoming friends and family who are visiting their loved one, arranging personalised keepsakes, and of course organising the funeral including the cremation or burial.

As a user researcher, I help my team learn about our Funeralcare colleague’s roles; which systems are already in place and where digital can make things better. The best way for me to do this is by visiting funeral homes, listening to my colleagues who work there, and watching them at work.

When I’m back in the office I relay what I’ve found out to the rest of my team. That said, I believe that ‘user research is a team sport’ so I always encourage my colleagues to come and see for themselves too.

Contextual research (actually going to the place to see how things work) makes sure we build a service based on needs rather than on our assumptions of what our colleagues need. So in this sense, my role is to make sure that when my team starts iterating on the service or building a new feature, we know it’s going to be useful as well as useable. Contextual research means that both time and money are spent wisely. We’ll iterate what we build along the way, of course.

I feel privileged to be on this team. Our Funeralcare colleagues only have one chance to arrange a loved one’s funeral. Building something that helps them do that smoothly and sensitively feels like a really important thing to work on and get right.

Mark Branigan
User researcher, Funeralcare