Which questions will my user research answer?

Why solve this user problem?

You want to find the biggest opportunity to improve your metric. So you’ll want to be confident that you’ve discovered the right user problems. (Which is why this phase is also called ‘Discovery’.)

You want to check:

  • Do we understand the problem?
  • Is it a big problem?
  • Is it a real problem?

You want to avoid:

  • Problems that users rarely experience
  • Problems that aren’t severe
  • Problems that don’t even exist

If you’re confident that you have already discovered the problem(s) to solve, you can move on. If not, then some user research might be useful.

When you do this ineffectively

When you do this effectively

  • You’ll be able to answer the question ‘Why are you solving this user problem? Why not solve some other problem?‘.
    • Specifically, this will guide what you do in future.
      • You’ll come up with ideas to solve this problem. (Rather than what sounds cool.)
      • You’ll assess your solutions according to how well they solve the problem.
  • Better still, users will want what you build and will jump through hoops to get it.

Easy mistakes to make

  • ‘Discovering’ a small problem because you were already looking for it. And missing the bigger problem that your research excluded.
  • Discovering ‘problems’ that don’t actually exist. It’s easy to believe that that users wake up in the morning and want your product. Generally, they don’t. They want to get some job done and are looking for some way to achieve that. (See https://hbswk.hbs.edu/item/what-customers-want-from-your-products .)
    • Easy way to spot this: stories about users that actually centre on your product or service. For example, “Users are looking for exciting new ways to engage with our weather app”. Or “Users want convenient ways to get discounts on their lunch”.

A way to better understand user motivations is to understand what they’re already trying to achieve. You can do this by finding out what they have spent time in the past struggling to do. For example, “Bill has tried to enjoy nature by going on camping holidays twice in the last year, but been disappointed both times.”

What improvement can you make to solve their problem?

You know which user problems you’re solving. Now you want to find out what you can do to solve them.

When you do this well, you’ll discover what the user needs to better solve the problem.

Here are some bits of jargon that describe what you’re looking for:

  • Use cases
  • Pain points
  • Moments of delight

You’re looking for ways to solve the problem. Again, if you already know would make it easier to solve the problem then move on.

How to understand solutions

A common error here is to ask users whether your solution would work. For example, ‘Would a social network for people who go camping make your holidays better?‘. You’re giving them the solution.

Instead, you’ll learn more about future behaviour by finding out what unmet need they have. For example:

  • ‘I try to share photos from my phone so that my friends and family can see. But they look bad and it takes a long time.’
  • Instagram found solutions to both of these existing, real frustrations. They applied filters and uploaded photos in the background.
  • No user started their day thinking ‘Wow I would love to use filters’. They thought ‘I want my photos to look good’. (Source: https://rocketship.fm/episodes/product_journeys_instagram .)

Search for ‘SMS’ here for another example: https://www.joincolossus.com/episodes/95054383/ravasio-supply-demand-and-the-outdoors?tab=transcript .

Asking users for ideas about how to solve the problem is an equally bad idea.

Another error is to think too narrowly about the available solutions. Most problems can be solved a number of different ways. (If there aren’t multiple solutions, then chances are your ‘problem’ is actually a solution. “As a user I want my discount code to be applied in one click so that it’s applied quickly”.

How effectively will you solve their problem?

You’ve discovered the problem that you want to solve. You’ve identified what specifically can be done to help them solve it. Now you want to verify that when you ship your product improvement, your change will actually induce that change in human behaviour.

Predicting actual behaviour is hard.

Again, your existing level of confidence will influence how useful you’ll find user research.

  • Bug fixes: will definitely work.
  • Common design patterns: normally work, but may not in this context.
  • Experimental new feature: no idea!