Building the right thing, for the right people by using customer interviews

The key to building a successful software company is solving a real problem for real people. Finding out who these people are and if they even have the problem you want to help solve is a crucial part of it. How useful is a product or business if its not solving a problem for anyone? Everyday our team at Neo works towards building the right things, for the right people.

Its very common for people to start with a business idea or a product feature — Aha! My app or business is going to be awesome! This feature is going to kick ass! The problem is that the passion can quickly take over and before you know it tunnel vision has kicked in blinding you from reality.

From “Talking To Humans” by Giff Constable

This tunnel vision is often the result of making grand assumptions about your end users without ever actually talking to them. Whats even more toxic is thinking you have validated your product or service by talking to people in your network of friends, family, or co-workers. The tricky thing to do (and obvious once you’ve done this) is to actually talk to strangers who you assume will use your product or service. Don’t spend months building something only for your product to fall on deaf ears and then more time and energy iterating your product into oblivion trying to find product market fit.

Try taking an inverse approach to customer acquisition as you do to product. With acquisition you need to invest heavily in the early stages with manual customer development while at the same time your product should take as little investment as possible. As you learn from interviews and experiments you should begin to invest more time into product and less into acquisition. Coming up with interview questions is the first step. What is it that you want to learn? The goal should be to understand your customer/users problems and getting them to talk and tell stories is the most valuable part. You’ll want to avoid questions that give you short answers or questions that lead to your assumptions.

Often the best learnings come from side comments the person makes as they explain their process or problem. Listen tentatively and take great notes; the language people use to explain their problem is often the same language you should use to sell them a solution. You’ll want to treat these early conversations with care as building trust and confidence can lead to referrals and follow-up conversations / interviews.

“Often the best learnings come from side comments the person makes as they explain their process or problem.”

The hardest part of this process is actually finding possible customers to talk with. Interviews should be set up however you can aquire them, both offline and online, but when conducting should be face-to-face or over video conferencing.

Being live and in-person is crucial for relationship building and being able to capture candid feedback while using e-mail, surveys or chat is significantly less helpful. You should also be testing different ways to ask for someone’s time. I am a believer that this should appeal to their ego and the problem at hand. Bribing people with gift cards or other rewards only skews your results; if someone isn’t willing to give you 10–15 minutes of their time the problem is most likely insignificant.

I approach acquiring these first customer interviews in a bit of an order:

1. Conduct 2–5 test interviews within your network. These should be mostly to practice your conversation/interview skills and make iterations. I don’t count these towards any real learnings.
2. Leverage Meetup.com to attend any and all meetups related to your market. If the product/brand is already developed, you can ask the organizers to speak briefly around the problem, otherwise you have to be on your hustle in the crowd. Ask for Twitter handles or emails and follow-up later regarding specifics for interview times.
3. Leverage Twitter.com to reach out to possible customers in bulk (but avoid more than 5–10 in a day as its easy to seem spammy). I suggest mitigating the risk of being spammy by using a few Twitter handles. Use WeFollow, MangeFlitter or other tools to find users in your market and Twitter Lists to find more of them.
4. Ask for referrals at the end of every interview. This is as simple as “Do you know anyone else we could talk to?” The warm lead of a referral makes it easier to convince someone to give you time and often their responses are of higher quality. Do anything you can to make this easier on the interviewee such as sending them a blurb that they can copy and paste.

I would emphasize the importance of approaching these possible customers with a long term relationship in mind. Some of my products biggest proponents and most helpful users have been people I engaged in 1:1 interviews with. Building trust early helps build confidence that you can solve their problem.

Compile feedback and rank their frequency.

I also find it helpful to compile feedback by extracting key concepts from interviews and sticking them in a google spreadsheet, then counting their frequency. This can make it a lot easier to quantify your interview data and as you approach 20 interviews can make it a lot easier

You’ll want to do just enough interviews that you begin to understand patterns in responses. If you can predict answers that is a good sign to either stop or make improvements to your questions so that you can learn more. Usually between 15–20 interviews is a good number but varies between problem, market and solutions.

Okay, awesome, you’ve found some people who have a problem you want to solve, but thats not enough. You want to quantitate data you can predict with precision to back everything up. To accomplish this second set of data points you will want to send traffic to a landing page that tests your value proposition based on a KPI. I cover scaling your learnings with a landing page in my next post you can read here.

I would love to hear your thoughts around validated product design. Feel free to connect with me on Twitter @realscottmcleod