Separating Change-Aversion from Real Problems in a Redesign
After a redesign, split 'upset because it changed' from 'upset because it's worse' before you react.
- Difficulty
- Advanced
- Time to result
- ~weeks to results
- Steps
- 4
- Confidence
- 92%
A redesign is fundamentally harder than a new feature because participation isn't voluntary — everyone is forced into the change. That guarantees two kinds of angry feedback: users upset merely because their habits broke (even if the new design is objectively better), and users upset because you genuinely made it worse. Reacting well requires telling these apart, primarily by comparing new user cohorts against long-tenured ones.
Origin
Söderström's framework for Spotify's 2023 home redesign, prepped with his teams before launch and referencing the historically contested Facebook News Feed change.
Core principles
- 01New features are voluntary; a redesign is imposed, so it carries a cost even for people who dislike it
- 02Two feedback types coexist: 'right but habits broken' and 'genuinely wrong'
- 03Objective data that the new design is better won't stop habit-based upset
- 04New-user cohorts lack the old habit and reveal the design's true quality
- 05Angry voices contain real signal once you filter the emotion
How to run it
- 1
Pre-warn the team it will hurt
Before launch, tell the team the odds of getting a big redesign exactly right are low, expect a wave of complaints, and that this is the nature of imposed change — not a sign of failure.
Pro tip Framing the pain in advance makes the team resilient enough to listen instead of getting defensive.
Watch out Redesigns can't always be A/B tested pre-launch when press attention forces a public reveal, so you learn after shipping.
- 2
Read past the anger to the actual complaint
Strip the hostile tone from feedback and identify what users literally cannot do anymore. In Spotify's case the complaint wasn't 'too much discovery' but 'I can't find my playlist.'
Pro tip Corroborate qualitative complaints with quantitative traffic shifts — users fleeing to search and library signaled they couldn't recall known items.
Watch out Some users express the real problem poorly; don't take the loudest phrasing at face value.
- 3
Compare new vs. old user cohorts
Look at users who never had the old habit. If new cohorts thrive where tenured users complain, the upset is habit-breakage, not a worse design. If both struggle, the design is genuinely worse.
Pro tip The physical-desk analogy: rearranging someone's 12-year-old desk upsets them even if the new layout is measurably better, because they were effective in the old one.
- 4
Hold beliefs at 100% until the data overturns them, then switch fully
Commit completely to a hypothesis, but the moment data disproves it, drop it and commit completely to the next. Stay unemotional and never get precious about work you invested in.
Pro tip This is 'strong opinions loosely held' — easy to say, hard to do, because people distrust those who change their minds.
Watch out Getting precious about a heavily-invested design is the biggest risk; you must be brutal about killing it when data says no.
In the wild
After flipping home toward discovery, Spotify faced angry tweets. Rather than reverting reflexively, the team read past the anger (users couldn't recall known playlists), confirmed it in traffic shifting to search/library, and recognized recall was something they did better than competitors — worth keeping.
→ They updated the hypothesis: keep strong recall, make discovery tools available but voluntary, pursuing the same goal via a better design.
Söderström cites the Facebook News Feed launch — intensely disliked at first — which nonetheless solved a real problem: users no longer had to run around Facebook collecting events themselves.
→ An initially reviled imposed change proved to be a genuine improvement, illustrating 'right but habits broken.'
Common mistakes
Reverting a redesign the moment people complain
Some anger is pure habit-breakage over an objectively better design; reflexive reversal throws away a real improvement (the Facebook News Feed would have been killed by this logic).
Getting precious about a heavily-invested design
Sunk effort tempts teams to defend a design after data disproves it; failing to switch fully to a new hypothesis wastes the learning the test was meant to produce.
Is it for you?
Best for
Product leaders shipping a forced redesign of a mature, heavily-used surface who must interpret a backlash correctly.
Not ideal for
Voluntary, opt-in new features where users who dislike it are simply unaffected and no imposed-change cost exists.
From the transcript
“one is you did something and it was right but people are upset because you changed stuff the other is you did something and it…”
“to look at new user cohorts that don't have that behavior versus old user cohorts”
“you have your pencil over here you have your notebook over there and I come in and I just rearrange all of it”
“You have to believe in things 100% until the data says no. And then you believe in something else 100%.”
From the episode
Lessons from scaling Spotify: The science of product, taking risky bets, and how AI is already impacting the future of music
Gustav Söderström (Co-President, CPO, and CTO at Spotify)