5 steps to a hypothesis-driven design process

Hypothesis design, as already the name may suggest is the process where you design solutions and test them with your users. It’s not a new technique but as working on assumptions may seem only intuitive and not factual one needs to be careful while proceeding with this type of process. 

You may ask why and when do people do it?

Let’s say your product manager gives you an ambitious project with high-end goals. The situation may not ideally have a straightforward solution. This problem can arouse a lot of questions and apprehensions in your mind. So you start with a hypothesis idea that will help you find your way through the unknown waters. And at the end of the process, you can come out with actionable steps.

What is a hypothesis?

Now considering what we just learned, it may sound like an assumption that in fact is true. An assumption becomes a hypothesis when there is a proposed solution and an outcome attached to it. The outcome needs to be tested too. In terms of design, we can consider a situation like changing the button and expecting an outcome like a better response rate. This hypothesis should be testable and verifiable. At the end of the test better user needs should be met and that should be the main goal. The tests also throw light on the pitfalls of the hypothesis. And will ultimately tell you how well you understand your customer’s behavior.

Benefits of the hypothesis-driven design process

• Less paper work-hypothesis design can save a lot of paperwork as anybody can share the hypothesis. This means that the entire team can pitch in their hypothesis and everybody will be on the same page.
• Informed prioritized roadmap-This means your roadmap will be based on evidence gathered along the testing.
• Constant learning-The testing can continue by making different assumptions and a lot of insight can be drawn.
• Customer centric-As already stated this technique will give you a solution that is indeed customer-oriented.

Step 1- Be curious and ask questions!

Just like a question about the fall of an apple helped Newton, it could you too. As a designer, you may be habitual in finding solutions to every problem in your project. But you need to pause a bit here and focus. Assumptions can only be hypothesized when you ask the right questions and can then provide solutions to them logically. It is equally important only to focus on the right problems. Hence you can start from scratch, start writing down all the questions that you have in your mind along with the assumptions. Building a product with numerous unanswered questions will make it harder. You can even turn your assumptions into questions. Try forming open-ended questions without answering them prematurely. It is also a good practice to have diverse minds brainstorming with you with questions. This will give you a wider perspective and a more productive session.

Another useful tip here could be asking existing customers. It may be difficult for them to understand the process but these can be a vital source as they are experiencing the problem firsthand. Try to ask them about the problems they are facing and what solutions they can suggest.

Step2-prioritize the questions

Once you are done noting down all the appropriate questions you need to plan and organize your questions. You can start marking them on your own or you can take help. First, start by thinking about which question would impact your users the most and prioritize it. Be realistic while making these assumptions in terms of the time frame under which you are operating.

Step 3-Turn assumptions into hypothesis

Once you have found out which questions should be prioritized you need to start turning them into a hypothesis. As already explained for these assumptions to turn into hypotheses we need to carry out extensive testing. Research needs to be given thought as there are multiple ways of testing these assumptions but ultimately they should depend on the time frame and scope. There are several ways to test, these methods should ideally be done qualitatively and quantitatively. And each assumption should be tested. You also need to understand what feedback would be needed to accept or disapprove of your assumption. These tests could be-

• Guerrilla testing-You can ask random strangers say in a coffee shop to share their opinions with you in exchange of a burger. Trust me this is the cheapest way.
• Face-to-face testing-Ask your consumers right away.
• Online testing-Several tools are available online that can give you quick results thanks to connectivity over the internet.
• Card Sorting-This helps you form the app or site structures and discuss the labeling with users.
• A/B testing-Can work well for design, wording, and colors.
• Click flow tests-This is needed to understand how well and whether people understand the processes. You can take help from the sign-up flow or purchase funnel.
• 5 seconds test-There are websites such as usabilityhub.com to get the users’ initial reaction to a page or a piece of information.

Step 4 – Measure the outcomes

After the interesting user tests that we just discussed it is time to focus on the results collected. These should be scientific, empirical, and measurable. This will enable you to obtain new knowledge. Remember you need measurable outcomes to deduce whether it was a success or a failure. Other tests involving page validation or usability testing can also be done. Additionally, you can write down what outcome would define success or failure. A well-defined experiment will be able to validate/invalidate your hypothesis. You can define a cut-off percentage for the test to pass or do a true or false outcome. While handing over the test it isn’t necessary to have each detail ready with you. Only what needs to be done and tested needs to be ready. This will save you from fretting about the non-vital aspects. You can also add more details as your finding grow and you can increase the scope of the research. After two or three experiments you would be ready to present the idea to your development team.

5. Step 5 – Learn and build

Not every test and experiment that you carry out will result in success. But the beauty of research is that it always gives you a better insight and educates you on something or the other. Remember that design is a never-ending process and each experiment will give you evidence for the next test that you carry. If you are successful good congratulations, you can roll out your new feature.

Concluding words and Key takeaways

Keep it simple-Try to avoid excessive documentation. This is to save you the trouble of updates, in case the hypothesis is rejected. Try to keep it as easy as changing the cards on your walls.

Be creative- There are always two ways to solve a problem one is the routine method and one with a new idea. While accessing the outcomes, it is essential to let the creative juices keep flowing as it may give you an unexpected result.

Don’t be easy on research-Users may not always know what they want yet a user-researched product is distinguishable from one that has had no user insight.