It’s frustrating to look at your metrics and wonder why you have high dropoff, low conversion, or low engagement.

Ultimately, there’s only so much digging you can do in Google Analytics or a spreadsheet before you need to try something different.

Google Analytics will tell you that people are doing things. But it will never tell you why. It will never tell you their motivations, their frustrations, and what they’re trying to do. But interviewing a subset of your users — though it may seem laborious at first — will give you the answers.

I’ve been interviewing users as a way to fix things that aren’t working and find new product opportunities for several years, and in my conversations with others, have found that it can be intimidating to get started.

1. What are you trying to solve?

This seems like an obvious step, but it’s important. You want to focus in on the problem you’re trying to solve as it relates to actions your users are (or aren’t) taking. “Why is our revenue lower than our goal” is a bit broad; instead, try to have a focused problem, such as “Why aren’t people upgrading from our free trial?”

Depending on how large your operation is, you might be able to widdle this down further — such as “Why aren’t people, who arrive on our site from Facebook ads, upgrading from our free trial?”

2. Recruit users

Who: Once you’ve determined the problem, you want to find users who are — and aren’t — taking that action. Focusing on the positives and the negatives means you’ll get a full picture of what’s going on.

For example, if you’re trying to determine why your free trial conversions aren’t higher, talk to the people who are converting, and the ones who aren’t. This will tell you how their motivations differ — which in turn can help you refine your messaging so you can find the most motivated users.

How: Email is an effective way to recruit users, but you can expect a low response rate — less than 5%, in my experience. When recruiting users, you want to hit several things:

1. Congratulate. “I noticed you signed up for a free trial of WidgetMaster. High five!”
2. Ask. “I’m looking for ways we can make WidgetMaster more useful. Would you have time for a 20 minute call sometime this week?”
3. Incentivize. “If you’re willing to take the time to talk to me, I’ll extend your free trial by a week.” Sometimes this isn’t necessary in B2B since the ability to make feature requests is a valuable currency, but it can he helpful to add if you’re getting trouble with response rates. In my experience, it is basically required in B2C.

How many? I generally find that 4–7 interviews is enough to give you the information you need to start testing changes. For the interviews, you should budget 30 minutes each…but leave an hour in case they want to go longer. If I’m working on a discrete question, I like to schedule the interviews all for the same week to keep my thoughts organized.

3. Interview!

Before starting the interview, I suggest having a rough script you’re looking to follow. There are a few simple things you’re looking to determine in the interview:

1. What motivated them to sign up for your product?
2. Before they signed up for your product, what were they using to solve their problem instead?
3. What’s the big goal they’re trying to accomplish overall?
4. Where are they struggling?

I generally focus the meat of the interview on those questions. Just as valuable, however, can be the reaching-for-the-door-question — “Well that about covers what I wanted to ask. Is there anything else you think I should know?” I’ve found that interviews can go on for another half hour after this question. It encourages people to open up to you and really dig into whatever it is they’re trying to do and why they haven’t accomplished it yet.

I’ve found that asking specific questions about the action you’re wondering about can be hit-or-miss. Asking “Why didn’t you convert to a paid plan?” is tempting, but I find that I will often get a better answer if I dig into their motivations.

During the interview, you’ll want to take notes. It can also be helpful to record the interview, but only if you’re planning to get it transcribed. If you’re a one-person interviewer with other responsibilities, you might find it easier to just take notes.

You’re done! Time to analyze and start testing

Once you’ve completed your interviews, the metal can really hit the road on your product. Now that you know more about your users’ motivations, maybe it turns out you just needed a simple messaging tweak — or you need to rethink an entire channel or flow.

Regardless of what you find, remember that interviews are an ongoing tool. Things stop working? Spreadsheet confounding you again? Get out there and talk to people. I guarantee you’ll learn new things about your customers and find something actionable.