fbpx
Menu
Designing

Tips on Designing Test for Product Discovery

In our previous product discovery article, we discussed about a number of test that are crucial in product discovery. These tests should be explored whenever you are in the process of validating problems and target customers. 

Here are a few things to consider when designing your tests:

1. Eliminate Bias

I can’t emphasize enough how important it is to have as objective a test as possible. This means not asking your friends, co-workers or parents to participate. Find total strangers who can give you honest and authentic feedback.

2. The Rule of 5

If you keep your criteria very tight — who you are asking and very specific things you are testing — you need not do more than five tests before you know where you are trending. But limit your variables per test (see next bullet).

3. Limiting Variables

The Rule of 5 only works if every test you do is limited to a couple of key questions you want answered. The more variables in a test, the harder it will be to discern what influenced an outcome. For example, if you are trying to test whether women ages 18–20 vs. women 30–35 have a problem finding a great yoga class, design a test that is the same in every way, just test it with five of each of these two different audiences. Similarly, limit variables in prototype tests such as in the smoothie test noted above, where when the founder tested add-ins at one particular site, that was the only variable he changed; all other aspects of the test remained the same including the site itself.

4. Breadth of Demographics

You may be designing a product that you believe everyone in the world will need OR that you believe only one target audience needs. Gender, income level, geography, etc. may or may not have an impact on adoption but you won’t know that until you parse things out early on and test a few. How a 13 year-old uses a product may be completely different than a 45 year-old (Facebook is a great example of this). Also, if you don’t test different demographics, you may miss an audience that could be in most need of your product.

5. Measured Outcomes

Start with a hypothesis of what will happen per test; ideally in measurable outcomes such as % of people who accept a restaurant recommendation or number of smoothie customers who want an add-in vs. those who do not.

Decide what you think success looks like for these tests. If your outcomes vary, then consider whether your test was valid and/or whether the learning lends itself to further testing or abandonment of an idea. In the case of the smoothie, the founder hypothesized that his target customer would want 5–6 flavor combinations, but found only 2–3 flavor combinations were most popular, thus he limited the flavor options in his MVP.

6. Leverage Existing Technology

Finally, in today’s highly tech enabled world, there are a number of ways to engage your target customers using what’s already out there to your advantage before building anything yourself;

  • Typeforms, google forms, etc. can capture form data.
  • Use Venmo to simulate online payment.
  • Texting can simulate alerts and notifications.
  • High fidelity web prototypes from Figma, Sketch, Invision, etc.
  • 3D printed mockups & scrappy hand crafted prototypes made from supplies you can buy online.

For this and much more Information relating to employability skills, employment opportunities, career advancement and entrepreneurship development; Join our Telegram and WhatsApp groups, and also follow us on Twitter and Facebook.

Culled from: Medium

Related Posts