Website Speed and Customer Confidence

The 400-Millisecond Window

In 1968, the computer scientist Robert Miller published a paper on human-computer interaction. Nielsen Norman Group’s Usability Engineering (1993) draws on that work to set out three thresholds for how people perceive time passing, and it is still the clearest account we have. Four hundred milliseconds is our own working line for a page that answers instantly, not a threshold the research sets; the three limits below are where the research says attention actually breaks:

1. 0.1 second (100ms): the threshold of instantaneous cause and effect. An action that lands this fast feels like direct, unmediated physical reality, the way a piano key does.
2. 1.0 second (1,000ms): the limit of unbroken thought. The delay becomes noticeable, but the person’s concentration survives it.
3. 10.0 seconds (10,000ms): the end of human patience. The train of thought is gone and the attention has moved somewhere else.

┌─────────────────────────────────────┬────────────────────────────────────────┐
│                 THE THREE THRESHOLDS OF PERCEIVED RESPONSE                   │
├─────────────────────────────────────┼────────────────────────────────────────┤
│ WHAT THE PERSON EXPERIENCES         │ HOW LONG IT FEELS                      │
├─────────────────────────────────────┼────────────────────────────────────────┤
│ Tap a button, the answer comes back │ 0.1s (100ms): instant                  │
│ The page has visibly started moving │ 1.0s (1,000ms): noticed, tolerated     │
│ They have left the tab entirely     │ 10.0s (10,000ms): gone                 │
├─────────────────────────────────────┼────────────────────────────────────────┤
│ INSTANT                             │ ABANDONED                              │
└─────────────────────────────────────┴────────────────────────────────────────┘

When your website takes three or four seconds to open on the phone in someone’s hand, it is not merely an inconvenience. It is a broken thread. The person came from a Google search, or an ad, or a message from a friend, and in four seconds they have already decided something about you.

They decided it without knowing they decided it, and they will not tell you. What they took from those seconds was confidence: that somebody is running this business carefully, and that the order they are about to place will be handled the same way.

1. Speed Is a Signal You Cannot Fake

In evolutionary biology, animals read power through signals that are expensive to fake: the run of a cheetah, the antlers of an elk, the flight of a falcon. The signal works precisely because it costs something.

Business works the same way, and a small business owner is often on the wrong side of this.

• Anyone can buy a cheap WordPress theme and paste a logo on it.
• Anyone can write copy claiming to be the best in the city.
• But almost nobody can make a site that opens instantly on a mid-range phone, holds its layout still while the images arrive, and still responds when someone taps the menu.

When a visitor experiences that, the inference they make is not technical. It is personal.

1. This opens faster than the app I use for banking.
2. Somebody paid attention to the small things.
3. A business that behaves like this will probably behave like this with my order.

Speed gets past the part of the mind that is looking for reasons to doubt you, before that part has finished getting ready.

2. The Feel of Something Made Well

You already know what this is like in the physical world. Walk into two workshops and you can tell within a minute which one keeps its tools properly.

One of them is a jumble. The drawers do not close. The bench is covered in parts of things. You find yourself moving carefully, as though something might fall over.

The other one has a place for everything. The drawers slide. The bench is clear. You can pick up the tool you need without asking.

The second workshop is not more expensive. It is the same work, organised so that the person using it is not slowed down. But it feels unmistakably better to be in, and you would say so if asked why.

A website is the same. A slow site makes the visitor move carefully. A fast site lets them move normally.

┌─────────────────────────────────────┬────────────────────────────────────────┐
│                        THE FEEL OF A WELL-MADE THING                         │
├─────────────────────────────────────┼────────────────────────────────────────┤
│ A DOOR THAT NEEDS A SHOVE           │ A DOOR THAT OPENS CLEANLY              │
├─────────────────────────────────────┼────────────────────────────────────────┤
│ Heavy, loose, a rattle in the frame │ Light, oiled, one smooth pull          │
│ You hesitate before using it        │ You do not think about it at all       │
│ A small sign the place is run down  │  A small sign the place is looked after│
├─────────────────────────────────────┼────────────────────────────────────────┤
│ FRAGILE                             │ TRUSTED                                │
└─────────────────────────────────────┴────────────────────────────────────────┘

Nobody rings a workshop to ask how its doors were hung. They just decide, quietly, whether the work in there will be good.

3. What a Fraction of a Second Is Worth

Google and Deloitte published Milliseconds Make Millions in 2020, and web.dev’s summary of the study describes 37 leading European and American brand sites, with data collected across over 30 million user sessions. The finding that matters most to you is the smallest one:

A 100-millisecond improvement in mobile site speed was associated with an 8.4% increase in retail conversions and a 9.2% increase in average order value, while mobile bounce rates on those same sites fell by 0.6% on the homepage, 5.7% on product listing pages and 1.9% on product detail pages.

In the same report, Google’s own research on mobile abandonment shows the probability of a visitor bouncing rising sharply as load time stretched: 32% more likely between one and three seconds, 90% more likely between one and five seconds, and 123% more likely between one and ten seconds.

Here is what that looks like on a small business budget. This is an illustrative model, not a client result:

┌────────────────────────┬──────────────────────────┬───────────────────────────┐
│         ONE MONTH OF ENQUIRIES (ILLUSTRATIVE, NOT A CLIENT RESULT)            │
├────────────────────────┼──────────────────────────┼───────────────────────────┤
│ THE MONTH              │ LOADS IN 3.5s            │ LOADS IN 0.6s             │
├────────────────────────┼──────────────────────────┼───────────────────────────┤
│ Ad spend               │ $600                     │ $600                      │
│ Clicks at $6 each      │ 100                      │ 100                       │
│ People who wait        │ 46                       │ 81                        │
│ People who enquire     │ 1                        │ 2                         │
│ Cost per enquiry       │ $600                     │ $300                      │
└────────────────────────┴──────────────────────────┴───────────────────────────┘

Same budget. Same clicks. Same copy. The difference is how many people were still there when the page arrived.

If you spend $600 a month on ads, the difference between one enquiry and two is $300 of your own money, every month, for the life of the site.

Conclusion: Speed Is Part of the Answer

Do not leave website performance to be settled after launch, by whoever is left holding the project.

Speed is part of what you are selling. It is the first thing a visitor notices and the last thing they will forget.

Strip out the plugin that loads a font nobody asked for. Cut the hero image down until it is the weight it should be. Put the page on a host that is closer to your customers than it is to you. Get the first thing on screen in under a second, and keep it still while the rest arrives.

A site that answers immediately is making an argument about the work, before you get to make it yourself.

Want to see what a fast site looks like for a business your size? Model your project with the Estimator or look through our Selected Work.