This article was written by Taylor Wescoatt, one of Seedcamp’s Experts-in-Residence. Follow Taylor on Twitter @twescoatt.
This title isn’t entirely accurate I have to admit. It should be ‘Validate or Die Slowly!’. If you’re reading this, you’ve read The Lean Startup and other best practice, and you know that validating your ideas is critical to success. It is also the quickest way to figure out if you’re off track, so you can re-evaluate your options.
At Seedcamp, we work with lots of startups at a very early stage, so validation is super important and we do it a lot. It’s hard though, for a good reason. Customer Validation research is about systematically challenging your own beliefs. As a founder, our sense of self-worth is often woven in with our startup idea, so forcing ourselves to question that is painful. Until you get used to it, then it’s fun!
There is a huge amount of best practice and industry built up around validation techniques, and you can spend many thousands of pounds hiring people to help, but how do you do it yourself on the cheap? Before we dive in, let me make a few suggestions;
- Be clear what you intend to learn from each engagement. “Meh” is the worst outcome.
- Be as brief as possible, avoid extraneous opportunistic questions. It weakens learning.
- Plan to test everything.
Here are the techniques I recommend;
Needs Analysis
- Simple Interviews: Asking three questions in a social setting, using the “Tell me about the last time you… X” format. Don’t talk about your product. For more info, The Mom Test.
- Surveys: Not much to say here, be careful about biasing your questions.
- Deep Interviews: Impression testing and task analysis work well here, as well as asking the user for their ideas on things. Two to three minutes of chitchat, 15 min of Q&A, leave 10 min for exploring.
Product Analysis
- Remote Testing: A really fast and cheap way to get a few unbiased qualitative perspectives on usability of live or demo pages. Usertesting.com, Trymyui.com.
- Guerilla Testing: Powerful and fun. Recruit for a small reward at the local mall or coffee shop, ask for feedback on different ideas.
- Live Testing: Some ‘user behaviours’ can only be tested in a live environment. A/B testing or closed-beta trials are good options.
Putting it all together, here are some places where each type of testing can be helpful in the drive towards Product Market Fit;
A couple of extra things to consider;
- Customer Councils can be a helpful format to do several types of research with.
- Give some thought to Dual Track Agile as a joint discovery/delivery methodology. It’s a great way to explore without over-committing your tech resources.
Building validation into your process is the way to go. At first, it may seem laborious, but you’ll be thankful that what you learn will have saved you a lot of time, you’ll be happier that what you build will get used, and you’ll have a stronger culture of caring about the customer within your team.
For a list of all my articles: http://seedcamp.com/eir-product-articles/