Sachin Agarwal’s blog

You're struggling with user acquisition.  No one's coming back.  Your product is robust and full-featured and you're adding new features every week.  You're only building the features that your users want, so you're not wasting any time on "wouldn't it be cool if" problems.  But no matter how hard you work, your graphs just aren't up and to the right. 

What to do?

One of the most valuable resources that startups (consumer web, again) have are the e-mail addresses of their registered but lapsed users.  It can be very valuable to reach out to them and figure out what part of your message or product didn't resonate.  Remember, if your "very disappointed" score is between 25% and 40%, it's often the message - and not the product - that isn't resonating.

Of course, if only 2 or 5% of your registered users remain engaged, it's probably both the product and the message. 

Here's a simple way to find out what's going on.  First, find a well-defined cohort of users.  This cohort could be from your last marketing push, or even "all users who signed up in winter" - you want they survey respondents to be as similar to each other as possible.  Grab the e-mail addresses for all of the ones who registered who fall out of your engaged metric (last 7 days, last 30 days, etc).  Send those lapsed users from that cohort the following e-mail:

That's it.

The survey should have these four questions:

• Why did you sign up for $product?
• What did you think you were getting with $product?
• How did $product disappoint you?
• What can we do to make $product better for you?

These are, of course, unstructured questions, so you're going to have to read all the answers.  Don't worry about missing stuff - the patterns will be repeated over and over again so you won't be able to forget about them.

If you want, you can add a checkmark box or a field to capture phone number for follow ups.  Make sure that you don't ask for name or force respondents to leave any personal information if they don't want to.  You want them to give their unvarnished feedback.

Take their feedback - figure out if it's the product or the messaging that's falling short.  If it's the product, add those features (and think about killing the ones they're not using).  If it's messaging, that's even easier to fix.