The IKEA effect
People value what they built themselves above what the same thing is objectively worth.
Labour invested in an object raises the builder's valuation of it, beyond what non-builders will pay. The inflation depends on completing the task; abandoned or failed builds do not produce it. It attaches to the specific thing the person made rather than to the category.
How it shows up in software
Configured dashboards, curated playlists, tagged libraries and customised workspaces all raise the cost of leaving, because the user built them. This is why onboarding that asks the user to set something up outperforms onboarding that sets it up for them, even when the automatic version produces a better result.
Using it well
- Let the user assemble their first working artefact during onboarding, even if you could generate a better one.
- Guarantee completion. A build the user cannot finish produces frustration rather than attachment.
- Make the thing they built prominent and editable, since the attachment is to that specific object.
- Offer a generated starting point they then modify, which gets a good result and still leaves their fingerprints on it.
Where it turns manipulative
- Adding assembly work with no functional purpose spends the user's time to manufacture switching cost, and users notice when setup is longer than the job.
- Designing configuration to be hard to export is lock-in built on a psychological effect, and the harm shows up the day someone tries to leave.
- Using the effect to justify shipping unfinished tools, on the grounds that users enjoy building, transfers your work onto them.
Where you have seen it
Notion
Users build their own page structures and databases rather than receiving a fixed app.
Spotify
User-made playlists sit alongside algorithmic ones and are the part people describe as theirs.
What the research says
- Norton, Mochon and Ariely, 2012Mixed evidence
Across experiments with folded origami, assembled storage boxes and built sets, participants bid more for items they had made than others would pay, and the premium disappeared when assembly was interrupted before completion.
Lab studies with small stakes. The completion boundary is the robust part; the size of the premium in real purchase settings is not established.
- Marsh, Kanngiesser and Hood, 2018Mixed evidence
Found the pattern in children as young as five, suggesting it is not a learned consumer behaviour.
Grades are a judgement about the evidence, not about how useful the idea is. Plenty of contested effects are still worth knowing, as long as you do not cite them as settled.