Liking
People say yes more readily to sources they like, and similarity is the fastest route to liking.
Liking is the tendency to comply with people who seem similar, familiar, attractive, or complimentary. It operates on thin evidence: a shared name, a shared hometown, a shared frustration. In products, the source is the brand voice, the founder, the support agent, or another user whose situation the reader recognises.
How it shows up in software
Liking travels through voice, through founders who write in public, through support replies that sound like a person, and through case studies featuring companies that look like the reader's own. Community-led products lean on it heavily, since the person recommending the tool is a peer rather than a vendor. It also decays: a voice that reads as charming at signup can read as smug at the third billing error.
Using it well
- Write case studies about companies that match the reader's size, industry, and stack, so recognition does the persuading.
- Let support agents sign with their own names and answer in their own words within a clear policy.
- Name the problem in the user's vocabulary, including the ugly internal terms they actually use.
- Keep the voice steady when things break, since a jokey error page after data loss destroys the affinity it was meant to build.
Where it turns manipulative
- Fabricating similarity, such as a sales tool that inserts a shared alma mater or hometown it scraped from a profile, manufactures rapport out of surveillance.
- Warm, first-name interface copy attached to a cancellation flow that hides the cancel button uses friendliness as cover for obstruction.
- Paying for testimonials from people presented as ordinary users, without disclosure, converts liking into a purchased signal.
Where you have seen it
Basecamp
Publishes opinionated writing from its founders, which makes buying the tool partly an endorsement of their view of work.
Linear
Writes changelogs and docs in the clipped vocabulary of the engineering teams it sells to, so the reader recognises their own register.
Duolingo
Gives lessons a consistent character voice, so the daily reminder feels like it comes from something with a personality rather than a scheduler.
What the research says
- Burger, Messian, Patel, del Prado and Anderson, 2004Mixed evidence
Incidental similarity, such as sharing a first name or a birthday, raised compliance with a request.
Lab studies with student samples and modest effects. The incidental-similarity literature has not been broadly replicated at scale, so treat the size with caution.
- Zajonc, 1968Well evidenced
Repeated exposure to a neutral stimulus increased liking for it, the mere exposure effect.
Bornstein's 1989 meta-analysis found the effect reliable across stimuli and cultures, though it flattens or reverses with excessive repetition and for stimuli that were disliked to begin with.
- Cialdini, 1984Mixed evidence
Describes the Tupperware party as a sales structure where the purchase decision is made about the host rather than the product.
Observational case material rather than a controlled test.
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.