Mental models for building products people love ft. Stewart Butterfield

Key ideas

Butterfield argues most product work should reduce the thinking users must do, and warns that growing companies drift into work that only looks like work.

From a machine transcript of the audio. Quotes link to the moment they are said.

1. 1

   Features pay off past a threshold

   Butterfield uses an S-shaped utility curve: early effort yields little, a threshold brings most of the value, and further investment stops mattering. He says teams often ship a feature below that threshold, add complexity and then conclude the idea was not worth doing. He adds that the bar keeps rising as people get used to better software, so basics like sign-up and password reset need to be revisited.

2. 2

   The real problem is usually comprehension

   Butterfield says removing friction matters when intent is high and specific, as with e-commerce checkout, but most visitors to a product like Slack barely want it and do not know what it is. He argues the job there is to make clear what the thing is and what to do next, and that counting clicks is the wrong goal, since many easy steps beat few fraught ones. He also says a product unlike its alternatives has to create its market, the point he connects to his 'We don't sell saddles here' memo.

3. 3

   Small courtesies can be your advantage

   Butterfield tells of watching people on a narrow Vancouver sidewalk fail to tilt their umbrellas, and of making 'tilting your umbrella is our opportunity' a saying at Slack. He says small conveniences such as magic links, notification defaults and a warning before messaging everyone built an emotional connection, and that this helped Slack spread as people moved between startups. Separately, he says he had staff chant that success would be measured by the value created for customers.

4. 4

   Growth breeds hyper-realistic work-like activities

   Butterfield says that as a company hires, the supply of work known to be valuable falls behind the people wanting work, and they fill the gap with activity that looks like work, without being stupid or evil. His example is Slack spending what he estimates as thousands of person-hours testing a tiny default in threads. He says the fix belongs to the leader, who must keep enough known valuable work available and say no up front, not scold the team.

5. 5

   Decide on pivots coldly

   Butterfield says the default advice is to persevere, but the real question is whether you have exhausted the realistic possibilities, as he says he had with the game Glitch. He argues distance is needed because admitting you were wrong to investors, staff and users is humiliating, and it can feel easier to let the company die slowly. He adds that changing an app six months in, with three people, hardly counts as a pivot, and flags survivorship bias in his own story.

In Stewart’s words

What’s available here

• Original source

  Linked. Opens on YouTube.

• Player on this page

  Available. YouTube’s own player, loaded only when you press play.

• Full text or transcript

  Not reviewed. No license to show a text version here.

• Our summary

  Draft. Written by us with AI from a machine transcript of the video’s audio (Whisper) and checked against it. Machine transcripts can mishear; each quote links to the moment so you can hear it. Not yet rights-reviewed.

• Ask questions with AI

  Not reviewed. Off until there is licensed text to work from.