March 20th, 2018

Working with the research funnel

“Our website needs redesigning. Our customers can’t find what they want on it.”

I think this is still a fairly typical client brief for a small to medium sized web design agency to receive, even in 2018. I’m also pretty sure you might hear something similar from time to time if you work in an in-house digital team. If you were approached by an internal stakeholder or client and this was your brief, how would you approach this job? What research would you do?

I have written previously about The Research Funnel and given some examples of the kind of research you might do during each stage but how does this really work in practice? I thought it might be helpful to think about it from the stand point of an example project.

Let’s take the example above and pretend this it’s a real project. Let’s say you are a freelance designer or a designer at a small agency and the client is a school. They have asked you to redesign their website because it doesn’t work very well and they are getting a lot of complaints from parents. You have already built some time for research into your costing. Excellent news! Where do you start?

Let’s remind ourselves what the funnel looks like:

The Research Funnel
The Research Funnel

This example project seems fairly typical of a redesign project and the first stage might be discovery. In terms of the research, you would probably need to do some strategic research. Some of the problems are known but the solutions aren’t. The audience groups are known but we don’t know much detail about each of them and what challenges we need to design for.

There are several ways you could approach this. In practice, this would depend on your budget for research but also how carefully you design your research at the outset. Here are a few ways you could approach this (using the funnel stages):

1. Linear progression from strategic to operational

1a. Strategic — we know where we are going

1b. Design solution

1c. Tactical

1d. Operational

This example is probably fairly typical of a redesign project. We know what we need to do, let’s not reinvent the wheel. We’ll talk to existing customers before we design and then after we have a solution. This was how many of our projects went at Mark Boulton Design. It’s certainly the safest option.

There is nothing wrong with this approach but in my experience, you might only design a solution that ‘just does the job’. Being lazy with your research — particularly during discovery, can sometimes lead to lazy solutions. So what else could you do?

2. Including adjacencies

Strategic — start with the known and work outwards

In my experience, this type of approach is less typical. Recruiting target or non core audience groups often requires more effort or money and that might result in those groups being missed out.

Competitor research is a no brainer as it mostly involves an extra investment in time or adding it into an existing user research interview but it’s surprising how often this isn’t done. Just by widening your net slightly and including a broader range of perspectives, you will have a richer source of data/insights.

You could then follow on with the other research stages described in the first example. This only gets a slightly broader range of perspectives. What else could you do?

3. Non linear

Exploratory and Strategic — start with the known and work outwards. Then figure out the broader ecosystem using non standard methodologies.



With this example, you build on the obvious and go beyond that. The focus for the exploratory stage is mostly on using non typical methodologies to enable you to gather a broader range of insights.

Clients and customers are often reluctant to go beyond a tight brief and will typically push back if you propose ‘extra’ pieces of research such as workshops or stakeholder interviews but you can always just add a few extra questions to an existing round of strategic interviews or widen out your recruitment criteria.

Instead of thinking of these as different exercises and even costing a project that way, it might be better to spend a bit of time designing a blended sample at the outset and including a few exploratory ‘big picture’ questions. Here is an example of how this could work in practice using a standard semi structured one to one interview. I have quickly drafted a sample discussion guide for our fictional project:

Example discussion guide
Example discussion guide

You can see that I have included exploratory, strategic and tactical questions in one interview. By following a similar structure for all your interviews, it makes it easier to analyse the insights. This structure could work for all of the audience groups we outlined in the third example.

For some groups who have less direct usage of the website (such as the LEA, other education providers, other schools), you would probably expand on the exploratory section and spend more time talking through those topics. For others, you might spend more time working through various tailored tasks. For example, can a teacher upload content easily, can a parent access photographs from a recent school trip and so on.

With limited budget for recruitment and very few researchers on the ground at my last job, this is how I approached many of the projects I worked on. In fact, I always include exploratory or strategic questions in the warm up section of any interview I am doing. Warming up a participant is so important. 

Even if I only have 30 mins with someone I will always spend 10 minutes just chatting and making them feel at ease. If you dive straight into tactical questions, you run the risk of missing out on a goldmine of rich insights that people often just mention casually in the bit of the conversation where they aren’t ‘testing’ anything or in analytical mode.

I also approach a user research interview as a facilitated conversation. I use a discussion guide but I don’t follow it too closely and use it as a structure to freestyle around rather than a script to follow. That’s a whole other blog post though.

So, next time you hear “our website needs redesigning” or you are asked to do some research, try to think about how you could broaden your sample and collect data and insights from different parts of the Research Funnel. Good luck!

Previously : Making room for my happy voice
Recently : What is Research Ops?