Your Website Has 50 Milliseconds to Look Trustworthy

In short: the aesthetic judgement of a page forms in 50 milliseconds and holds at 17 (Lindgaard et al. 2006; Tuch et al. 2012). Visual appeal governs fast rejection, information quality governs the decision to trust (Sillence et al. 2006). In online retail, site design drives acquisition while order fulfilment drives repurchase (Blut et al. 2015). And the most quoted brand-consistency figures come from surveys of stated intentions, which explain roughly 28% of actual behaviour (Sheeran 2002).

Anyone opening your website has already decided whether you look credible before reading a single line. The judgement of how the page looks forms in fifty milliseconds, and from that moment it colours everything the visitor reads next.

For someone about to sign a redesign quote, a different question matters. How much is that twentieth of a second worth against everything else, and when is new graphics the wrong place to spend.

How fast does someone judge your website?

Fifty milliseconds, a twentieth of a second. In 2006 Gitte Lindgaard and her team at the Human-Oriented Technology Lab of Carleton University in Ottawa showed homepages for half a second and then for fifty milliseconds: the appeal ratings from the two conditions matched (Behaviour & Information Technology, 25(2), 115-126).

Six years later a group that also worked for Google identified which two properties get read that fast. Alexandre Tuch and colleagues had 119 screenshots of real websites rated at shrinking exposure times, down to 17 milliseconds, and isolated visual complexity and prototypicality, meaning how closely a page resembles the layout a visitor expects for that category (International Journal of Human-Computer Studies, 70(11), 794-811). Very dense pages start behind. Pages with a recognisable structure start ahead.

The practical consequence lands on anyone who wants to stand out through page structure. A layout invented from scratch pays a toll at the door that the content then has to earn back, while the personality of the brand can live entirely in the colours, the photography and the writing without moving the place where a visitor expects to find the menu.

Does design matter more than what you write?

It matters first. In 2003 B.J. Fogg and his team at Stanford University asked 2,684 people to rate the credibility of live websites and coded the unprompted comments: visual design appeared in 46.1% of them, ahead of information structure and information focus (Proceedings of DUX 2003).

Three years later Liz Sillence and Pamela Briggs followed real people searching for health information and found a clean division of labour between the two (International Journal of Human-Computer Studies, 64(8), 697-713). Visual appeal predicted the fast rejection of a site. Information credibility and personalisation predicted the decision to trust it and come back.

These are two gates in sequence. A beautiful site with generic content clears the first and stalls at the second, while a site full of substance in shabby packaging never gets read at all. When an agency proposes to work only on the first gate, the quote covers half the problem.

Why does a polished site also look more reliable than it is?

Because the aesthetic judgement spreads to the other judgements. Noam Tractinsky had 132 students use cash machine interfaces with identical behaviour and different visual quality: perceived beauty and perceived usability correlated before and after use, and the visual level moved the final judgement while actual usability did not (Interacting with Computers, 13(2), 127-145).

It is a credit line, and sooner or later someone calls it in. The visitor uses appearance to estimate what cannot yet be verified, and keeps using it until better information arrives.

The meta-analysis by Yeolib Kim and Robert Peterson across 150 empirical studies shows that online trust grows out of perceived security, perceived privacy and perceived service quality, and that trust predicts repurchase (Journal of Interactive Marketing, 38, 44-54, 2017). Aesthetics open the credit line, substance repays it.

Where is investing in appearance worth it?

In online retail, design carries acquisition and hands over the lead as soon as the customer has bought. Markus Blut's meta-analysis across 89 studies and more than a hundred thousand people breaks digital service quality into four dimensions and finds order fulfilment to be the strongest driver of satisfaction (Journal of Retailing, 91(4), 679-700, 2015).

The same work reports that the weight of the four dimensions shifts with country and industry. In professional services and healthcare, where the buyer cannot verify quality before purchase, visible signals count for more than in a category where the product speaks for itself.

An online store with an acceptable conversion rate and weak repurchase has a problem no redesign fixes. Before signing, look at which of the three stages in the table is the broken one, and ask the agency to say it with the numbers already in front of them.

Do the brand-consistency numbers hold up?

Three figures turn up in almost every deck on this subject, and none of the three measures what it appears to measure.

The first two collect what people say they would do. Paschal Sheeran's review of 422 studies and 82,107 participants finds a correlation of 0.53 between stated intention and later behaviour, so intention explains about 28% of what people actually go on to do (European Review of Social Psychology, 12, 1-36, 2002). A survey of two hundred consumers measures far less than the thing the quote promises to move.

A proposal that opens with these three figures is passing round someone else's promotional material in place of evidence. The same looseness will come back in the monthly reports, and the same test applies when you assess the case studies you are shown.

Which levers have the strongest evidence?

1. Maximum legibility wherever the reader compares alternatives. Eleven preregistered experiments with 9,042 participants show that when material is hard to decode, people choosing between options lean less on the numbers and more on intuitive cues such as a familiar brand or a country of origin (Mohsenin and Munz, Journal of Consumer Research, 2025). Price lists, comparison pages, configurators and quotes want high contrast, prominent figures and no decorative background behind the text.
2. Specific proof placed next to the point where doubt appears. A client logo beside the project it belongs to, a number with the period it refers to, the name and role of the person quoted, a link to the page where the figure can be checked.
3. Visual assets held still for years. The research on distinctive brand assets, gathered by Jenni Romaniuk in Building Distinctive Brand Assets (2018), shows that recognition builds by accumulation and restarts from zero with every deep change. A brand redrawn every two years never reaches the point where the colour alone says who you are.
4. Security and privacy visible without asking. Real contact details, a company number, a readable policy, return terms findable in two clicks. In the Kim and Peterson meta-analysis these are direct antecedents of trust, and their absence costs more than their presence earns.

What to ask the agency proposing a redesign

1. Which number should move, and where do we read it today?
2. How much of the problem do you attribute to the page, and how much to what happens after the click?
3. What are you basing it on when you say the problem is the appearance rather than the clarity of the offer?
4. What stays identical about the brand after the work, and why?
5. Who measures before and after, with which tool, over which window?
6. If the number has not moved in six months, what happens to the contract?

A promise like "+120% conversions" deserves one more question. Stefano DellaVigna and Elizabeth Linos compared 126 controlled experiments covering 23 million people against what academic journals publish: interventions that work in the field average 1.4 percentage points of effect, against 8.7 points in published studies (Econometrica, 90(1), 81-116, 2022). Anyone quoting a single case with a three-digit number is showing the exception that worked, and the same caution applies when you assess an agency portfolio before signing.

Frequently asked questions

Can fifty milliseconds really lose you a customer?

They lose you the benefit of the doubt. The fast judgement raises the burden of proof the visitor asks for straight afterwards. On a considered purchase that burden can be met with clear content, on a passing visit it rarely is.

Rebuild the site or rewrite the content?

It depends where visitors stop. If people arrive and leave within seconds, the problem sits at the entrance. If they stay, read and never ask for a quote, the problem is the offer and the proof, and new graphics leave the number where it was. The cost of a rebrand is worth comparing against the cost of a content review before deciding.

How often should a logo and colours change?

As rarely as possible. Recognition of a visual asset builds through accumulated exposure, so every deep change sets the counter back. A change earns its keep on a brand damaged by a crisis, on a merger, on a change of ownership and on a position that has stopped describing what you sell.

How do you measure whether the redesign worked?

With a metric and a time window chosen before launch. The four that hold up are conversion rate by traffic source, qualified enquiries per month, cost per enquiry and the share of enquiries that turn into quotes. Compare against the same period last year, because seasonality alone can explain a movement that looks like the work of the new design. In the agency contract these lines belong before the first line of code.

Sources

1. Lindgaard, G., Fernandes, G., Dudek, C., Brown, J. (2006), Attention web designers: You have 50 milliseconds to make a good first impression!, Behaviour & Information Technology, 25(2), 115-126. DOI 10.1080/01449290500330448.
2. Tuch, A.N., Presslaber, E.E., Stöcklin, M., Opwis, K., Bargas-Avila, J.A. (2012), The role of visual complexity and prototypicality regarding first impression of websites, International Journal of Human-Computer Studies, 70(11), 794-811. DOI 10.1016/j.ijhcs.2012.06.003.
3. Fogg, B.J., Soohoo, C., Danielson, D.R., Marable, L., Stanford, J., Tauber, E.R. (2003), How do users evaluate the credibility of Web sites? A study with over 2,500 participants, Proceedings of DUX 2003. DOI 10.1145/997078.997097.
4. Sillence, E., Briggs, P., Harris, P.R., Fishwick, L. (2006), A framework for understanding trust factors in web-based health advice, International Journal of Human-Computer Studies, 64(8), 697-713.
5. Tractinsky, N., Katz, A.S., Ikar, D. (2000), What is beautiful is usable, Interacting with Computers, 13(2), 127-145. DOI 10.1016/S0953-5438(00)00031-X.
6. Blut, M., Chowdhry, N., Mittal, V., Brock, C. (2015), E-Service Quality: A Meta-Analytic Review, Journal of Retailing, 91(4), 679-700.
7. Kim, Y., Peterson, R.A. (2017), A Meta-analysis of Online Trust Relationships in E-commerce, Journal of Interactive Marketing, 38, 44-54. DOI 10.1016/j.intmar.2017.01.001.
8. Sheeran, P. (2002), Intention-Behavior Relations: A Conceptual and Empirical Review, European Review of Social Psychology, 12, 1-36. DOI 10.1080/14792772143000003.
9. Mohsenin, S., Munz, K. (2025), How Perceptual Disfluency Affects Consumer Choices, Journal of Consumer Research. DOI 10.1093/jcr/ucaf062.
10. DellaVigna, S., Linos, E. (2022), RCTs to Scale: Comprehensive Evidence from Two Nudge Units, Econometrica, 90(1), 81-116. DOI 10.3982/ECTA18709.
11. Romaniuk, J. (2018), Building Distinctive Brand Assets, Oxford University Press.

Note: the three images in this article are generated with artificial intelligence. The figures cited come from peer-reviewed studies and international meta-analyses, each checked against volume, pages and DOI; they are not proprietary data of Migliore Agenzia.