Trust formation
Trust is judged fast on surface cues, then revised slowly by behaviour over time.
Initial trust rests on cues a person can read in seconds: visual quality, familiarity, third-party signals and the presence of a plausible operator. Sustained trust rests on competence, predictability and honesty about limits. The two run on different clocks, and surface trust can be destroyed by one behavioural breach.
How it shows up in software
Users decide whether to enter a card number, connect an account or upload a document long before they have evidence. They read layout quality, spelling, whether prices are visible, whether a human company appears to exist, and whether the flow matches what they expected. After that, trust is maintained by doing the boring things reliably and by how the product behaves when something goes wrong.
Using it well
- Show the cost, the terms and the cancellation path before you ask for payment details, not after.
- Make the system's state legible at every step: what was saved, what was sent, what is still pending.
- Say what the product cannot do. Stated limits are one of the few cheap, honest trust signals left.
- Handle failure in your own voice. A specific, accountable error message builds more trust than a flawless week.
Where it turns manipulative
- Borrowing trust signals you have not earned: fake review counts, dormant security badges, logos of companies that merely trialled the product.
- Front-loading warmth and honesty in onboarding, then relocating the real terms into a settings page nobody opens.
- Using an established trust relationship to slip through a scope change, for example expanding data sharing in an update described as a bug fix.
Where you have seen it
Stripe Checkout
Shows the merchant name, the exact amount and the currency on the same screen as the card field, with no additional steps introduced after payment details are entered.
Wise
Displays the exchange rate, the fee and the recipient amount before the sender commits, rather than resolving the rate afterwards.
1Password
Publishes a security white paper describing the encryption model, including what the company itself cannot access.
What the research says
- Fogg, Soohoo, Danielson, Marable, Stanford and Tauber, 2003Mixed evidence
In a study of 2,684 participants evaluating websites, design look was the most frequently mentioned reason for a credibility judgement, ahead of information structure and company motive.
Participants commented on what they noticed, which is not the same as what drove the judgement. The finding is about stated reasons.
- Lindgaard, Fernandes, Dudek and Brown, 2006Mixed evidence
Visual appeal judgements of web pages made after 50 milliseconds correlated highly with judgements made after longer exposure.
This is about visual appeal, not trust. The common leap from 50ms appeal to 50ms trust is an overreach.
- McKnight, Choudhury and Kacmar, 2002Mixed evidence
Modelled and tested initial trust in a web vendor, separating disposition to trust, institution-based trust and beliefs about the specific vendor.
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.