The guide to creating how might we statements for design thinking
“How might we…”
One simple question. So many possibilities.
The “How might we” (HMW) statement is a tool designers frequently use to reframe problems into a question. By reframing the problem, it enables the design team to generate not just better ideas, but more of them as well.
For example, maybe at your most recent team retro, your team identified that it is taking way too much time to update your work management system. You might describe the problem as “It takes too long to update the system.”
If you just start from that problem statement and start trying to come up with ideas, you might get a few solutions but pretty quickly you may run out of creative energy. Why not try putting on a 1 minute timer now and writing as many ideas as you can for this fictional team?
Instead, we can reframe the problem by asking “How might we make it faster to keep the team up to date with progress?”
Suddenly the options for solutions feel a lot broader. Try putting on another 1 minute timer and see how many ideas you can generate this time.
As you can see, the “how might we” is a useful tool. And it’s not just for solving problems from your team retro. You can use it in your design processes to develop features or use it when spiking around an issue to come up with more potential solutions. You can even apply it to organisational design questions or developing your team or business strategy.
So how do we actually go about developing “how might we” statements and using them in design thinking? This guide takes you step by step through the process of creating how might we statements that lead to great ideas.
Step 1: Define the business or customer need
Before we turn our problems into “how might we” questions, we need to define what the problems are. That’s what steps 1 and 2 are all about.
Now, problems can be identified in a whole number of different ways, as I touched on in the intro. And there’s a whole process and many tools within the practice of design thinking that will help you identify clear and meaningful problems. In fact, you may be reading this guide because you’ve finished your problem research and need to move into solution creation.
However, if you’re running a rapid design or problem-solving process and you’re right at the start as you’re reading this, I recommend you make sure you are solving the right problem before you start using the “how might we statement”.
Step one is to ask what need is motivating you or your team to look for a new solution. This can be a business need such as increasing revenue or decreasing cost. It can also be a customer need such as a lower friction experience or more effective service. Get as specific as you can in describing this need.
Step 2: Identify the problem you are solving
Once you are clear on the need, you can describe the problem. To do this ask “what is blocking us from solving the business or customer need?” You might have a list of possible answers to this question.
Capture as many as you want and then use a simple tool to rank them in order of importance. If you’re working in a team, perhaps have the team vote for the most important. If you’re working on your own, a tool like the value-ease matrix can be really helpful.
Step 3: Reframe your problem to a how might we statement
Here is where we bring in the magic question. Pick the top problem you identified and turn it into a question. You can start by literally putting the words “how might we” at the start of the sentence describing the problem. From there, you may want to refine the language so that the “how might we” statement makes more sense.
It also helps to try phrasing your “how might we” in multiple ways. Try writing out two or three different ways of asking “how might we” about the problem you have chosen. Pick the one that you find most interesting or starts to prompt you to think of solutions.
Step 4: Test your how might we statement
A good “how might we” statement can be the difference between having many creative, interesting potential solutions, and having only one or two obvious ideas that have been tried before. So the first few times you or your team create a “how might we” statement, it is useful to test you have created an effective one.
Three issues I commonly observe when teams are writing “how might we” statements for the first time are:
- They put a solution in the statement and constrain the volume and variety of ideas they generate
- The question is too broad and teams are prompted to produce ideas that are too abstract to be actionable
- The question over-simplifies the problem and makes it difficult for teams to be creative in the types of solutions they suggest
Not to fear — I have a test for each of these scenarios to make sure you are working with a quality “how might we” statement. First, put on a 1 minute timer and attempt to write as many ideas as you can in that time. These do not need to be smart or viable ideas. It’s simply to test the volume and types of ideas you get from your “how might we” statement. Then go through each of these three tests:
Test A: Can you think of more than one solution that answers your “how might we” question?
If you struggle to think of more than one or two ideas immediately as you read your “how might we”, this is usually because your “how might we” contains a solution. If this is the case, try and remove any part of the “how might we” that is itself a solution.
For example, when working with digital product teams, I often notice teams have “how might we” statements like “How might we automate this step to make the process easier?” or “How might we use machine learning to make this faster?” In these statements, automation and machine learning are solutions that make a process easier or faster. But if you keep those words in your “how might we” you are constraining your team just to ideas involving automation or machine learning!
If this happens, remove the words that specify a solution and rewrite your “how might we”. In the examples from above, we might say “How might we reduce the number of steps the customer has to take?” or “How might we reduce the amount of manual processing required in this process?”
Test B: When you write ideas for your “how might we”, are those ideas things you can make or do?
The key to the second test is to look for action words in the ideas you’ve generated. If your ideas are vague or abstract, it might mean your “how might we” is too broad. Ideas like “a better user experience” or “more interesting content” are not very concrete or actionable.
If this happens, try to make your “how might we” more specific. It can help to go back to the need and problem you defined in steps 1 and 2 to look for the specifics of the problem we are trying to solve. Once you are getting ideas more like “remove the number of menu items to reduce clutter in the UI” or “create a video on topic X” then your ideas are actionable, suggesting you have a good “how might we”.
Test C: Do your ideas feel like they over-simplify the problem?
If you read back over your ideas and your brain starts racing with caveats and additional complexities that need to be considered before any of the ideas can be taken seriously, your “how might we” statement might be too narrow, or over-simplifying the problem.
If this is the case, look to broaden the “how might we” statement. This issue can be closely related to the issue of putting a solution into your “how might we” statement. For example, if your question is “how might we simplify the user interface” but the motivating need that you started with is something like “increase the number of website visitors that sign up” then you may have narrowed your problem space a little too much. Something like “how might we make it easier for users to sign up for our website” might be a more useful question to ask.
Final note on testing
You don’t need to run through all three of these tests every time you write a “how might we”. As you use “how might we” statements more often, you and your team will develop a natural eye for writing them.
However, it can be useful to keep these tests in your back pocket. On the occasion that your team is really struggling to generate ideas in response to a “how might we”, you can run the statement through these tests to make sure the issue isn’t how you’ve framed the problem.
Step 5: Start generating solutions!
The purpose of creating a “how might we” statement is to frame a problem in such a way that it is easier to generate solutions. So once you have a good “how might we” statement, write it up somewhere you or your team can see it easily and start generating ideas.
The “how might we” statement can be useful both in time-boxed ideation sessions such as a design thinking workshop, or as a tool that you use to guide ongoing idea generation and problem-solving that your team is working on. It just depends on how your team operates and how complex the problem you’re solving is.
If you’re new to creating “how might we” statements, or you’re introducing this approach to your team for the first time, I’ve put together the steps in an easy to use worksheet:
You can get a printable / editable copy in Google slides here.
The takeaway: writing a good “how might we” does take some practice. But a quality “how might we” can be the difference between generating limited, obvious solutions and finding innovative, interesting solutions. Where can you use “how might we” to solve a problem you’re working on?
