How do you decide what features to build (or kill)?

How do you decide what features to build (or kill)?
Share article:

Every Friday, we’re answering your questions about business, startups, customer success and more.

Happy Friday!

This week’s question is:

We spent the first six months of our development completely screwing this up.

We built the product that we thought people wanted, and when we considered features, we simply asked “does this add to the product?” The answer to this wrong question was, far too often, yes.

And we ended up launching with a bloated, complex product that simply confused people.

These are all of the pages from our original site, listing all of the features we stuffed in to the product:

Check out this post for more on that early fail.

Since then, we’ve taken a much more methodical approach to adding new features, and at the core of that approach is customer development.

I really can’t stress that enough: it is amazing how many of your problems and business questions can be solved by making a conscious effort to systematically talk to as many of your customers as possible, and listen very, very hard.

You’ll find clarity like you’ve never had before about your product roadmap, your marketing, your customer service… everything.

Ultimately, our customer development efforts are the biggest driver of our product decisions; we consider what we hear and from how many customers we hear it from.

We also take into account support tickets (we use labels in Groove to track how many customers are requesting various features) and other customer research (NPS surveys, for example.

Ultimately, it’s a combination of a mathematical decision and a gut one, but my biggest piece of advice here would be to arm yourself with as much insight into what your customers actually want as possible, because it’ll make all of your decisions better.

Alex Turnbull
Alex Turnbull Alex is the CEO & Founder of Groove. He loves to help other entrepreneurs build startups by sharing his own experiences from the trenches.