What product features should I focus on?

deschart

My favorite slide of the year came from Des Traynor, the co-founder of Intercom who spoke at our SCALE event, on October 23-24, 2014. At that event, he was the highest ranked out of 54 speakers.

This is extremely notable because:

a) He was also a sponsor — yes, a SPONSOR gave the best presentation at the event.

b) He was rated a 9.61 out of 10 by 900 founders of venture-backed companies.

So notable was his talk that I asked him to give it again to our LAUNCH Incubator and he agreed. The seven startup companies and their founders were blown away and asked dozens of questions. I literally had to kick them out of the (slick and new) Galvanize offices over on Tehama Street in SF-SOMA.

[ Click to Tweet (can edit before sending): http://ctt.ec/w23er ]

The Slide of the year

In the slide above you see a chart with two axes: the Y is how often people use a feature and the X is what percentage of people use a feature.

Des explains the chart:

“One way to think about this is this (he shows the slide): the core of your product is here (points to top right),  you make an improvement here everyone appreciates it and it pays off big time. You’re in Dangerville over here (points to top left), right? Small number of people heavily dependent on small PC or product. The reason I call this Dangerville is because it’s the road to consulting-ware” (as in software made by consultants for clients).

The top right of the slide, to be clear, is a feature that everyone uses all the time. If we were talking about Gmail it would be reading email, if we were talking about Tinder it would be swiping left and right, and if we were talking about Inside.com (or Reddit) it would be scrolling up and down the “top” feed.

Now, if you get to the top left of the chart, it’s a feature that a few people use all the time. This basically means it’s customizable software and — in my mind — should be done via your API.

If you try to fill the chart, Des points out, you’re making a huge mistake:

“You’re going to have a very messy product if loads of things are heavily dependent upon by very small groups of people and you’re putting it all together in one product. It means you’re on your way towards being a swiss-army knife — a small group of tangentially useful features, independently used. There’s no benefit to being complicated.”

Percentages of your Time

After thinking about Des’ talk I thought about what I should spend our time on at Inside.com, This Week in Startups, angel investing, and at the LAUNCH Festival. So, I put a little chart together of % of our effort and resources. 

Here is what I came up with: half the time on your main feature, 15% on each of your next two features and then ~7% on the next three. So, take your product, define the top six features, and think about how much time you’re putting into making each feature awesome — this might help.

For Reddit it might be 50% on the homepage experience, 15% on the subreddit feature, and 15% on the commenting systems. The ~7% on the other three features could be the submission process and other things that either “some people use all the time,” “all people use some of the time,” or “most people use most of the time.”

jasonsdeschart

In Summary

As a founder you should watch Des’ talk tonight, and tomorrow you should buy everyone at your company lunch and show these charts. Map your product against them and ask yourself:

a) are you putting enough attention towards the features “all of the people use all of the time”?

b) are you doing any features that “few of the people use very little of the time” — that you should kill / take out of your product (failed experiments causing overhead!)?

c) are you doing any consulting-like features that “few of the people use all the time” — and can a 3rd party build them? Also, are they taking up a lot of time? If not, no big deal; if yes, maybe you need to be paid a lot for them.

Here’s the video: 

Here’s where you can find Des:

https://twitter.com/destraynor
https://twitter.com/intercom
https://www.intercom.io/