The test

Open your homepage on a laptop. Get someone who does not work for you and has never seen it. Show them the screen for five seconds. Close the laptop.

Ask two questions:

  1. What does this business do?
  2. Would they serve someone like you?

That is the entire test. It takes under a minute and it is the most useful minute you will spend on your website this year, because it measures the only thing the first screen is for.

Most homepages fail it. Not because the writing is bad, but because the writing was done by someone who already knows the answer.

Why you cannot run this on yourself

You cannot un-know what your business does. When you look at your own headline, you are reading it with fifteen years of context loaded, and it makes perfect sense. A stranger is reading eleven words with nothing behind them.

This is why "Excellence in every detail since 1998" survives on so many sites. To the owner it is a summary of a career. To a visitor it is a sentence that would fit equally well on a jeweller, a law firm, a bakery, or a machine shop.

The test only works with a real stranger. Three is enough. If two of three cannot say what you do, the headline is the problem, and no amount of SEO work downstream will compensate — you can rank first and still lose the visit in five seconds.

Twelve words

Here is the constraint we use, offered as a working rule rather than a law: the main headline should be twelve words or fewer.

The reasoning is practical rather than mystical. Usability research going back decades finds that people scan web pages rather than read them, weighting the first words of a line heavily and dropping off fast. A headline past a dozen words stops being read as a sentence and starts being skimmed as a block — which means the words at the end, where people tend to put the actual point, are the ones that get lost.

More usefully: in practice, almost every headline that runs long turns out to contain two separate ideas. Splitting it into a headline and a subhead is not a compromise, it is the correct structure that was hiding inside the long version.

Nine words is better than twelve. Six is often better than nine.

Four patterns that pass

You do not need to invent a form. Four patterns cover nearly every small business, and all four survive the five-second test because none of them require prior context.

Service plus place. Emergency plumbing in Plano and North Dallas. The most boring option and the highest-performing one for local businesses. It answers both test questions in one line.

Outcome for a named audience. Bookkeeping for contractors who hate bookkeeping. Works when the audience is specific enough that the right person recognises themselves immediately. Fails when the audience is "everyone."

The specific claim. Same-day AC repair, seven days a week, Tarrant County. Use when you have one real differentiator that a competitor would struggle to copy. It has to be true and it has to be checkable.

The direct question. Need a new roof after the April hail? Good for seasonal or event-driven demand, where the visitor arrived with a very particular problem. Weak as a permanent headline, because the event passes.

Every one of these is plain. That is the feature.

What fails, and why

The abstract noun. Innovative solutions for the modern enterprise. This describes nothing. Cover the logo and it could belong to any company on earth — which means it is doing no work that your logo was not already doing.

The internal-vocabulary headline. Comprehensive residential drainage services. The customer with a clogged sink is not searching that phrase and does not think in it. The gap between how businesses describe themselves and how customers describe their problems shows up here before it shows up in your rankings.

The tagline in the headline slot. Building trust, one home at a time. Taglines are for people who already know you. A stranger from search does not have the relationship that makes a tagline mean something.

The clever one. Wordplay requires the reader to hold two meanings at once, which requires them to already understand the first. It is a second-visit pleasure being asked to do first-visit work.

The one that hides behind an image. Sometimes the headline is fine and the problem is that it is 24px white text over a busy photograph, on a hero that takes four seconds to load. If your visitor has not seen the headline yet, it does not matter what it says — which is why the hero image is worth treating as a performance problem.

The subhead is where the personality goes

The most common objection to plain headlines is that they are boring. That is a fair objection and the answer is structural, not stylistic: the headline is for identification, the subhead is for persuasion.

Emergency plumbing in Plano and North Dallas. Licensed, insured, and usually at your door within two hours. No call-out fee, ever.

The headline gets a stranger oriented in under a second. The subhead is where you sound like a person, make the specific promise, and handle the first objection. Give yourself one or two lines, no more — this is still above the fold, and the visitor has not agreed to read an essay yet.

Under that: one button that does the thing. Not four.

What "above the fold" means now

The phrase is borrowed from newspapers and it survives because the underlying idea does. There is a screenful of content that a visitor sees without doing anything, and there is everything else, which requires a decision.

People do scroll — that objection has been thoroughly answered by research, and there is no need to cram a whole site into one screen. But scrolling is a choice, and the first screen is where they make it. The job of the top of the page is not to sell. It is to earn the scroll.

What has to be there:

  • The headline, passing the five-second test.
  • A subhead that adds the specific promise.
  • One primary action.
  • Enough visual clarity that the eye knows where to go first — which usually means more white space than feels comfortable, not more content.

What does not have to be there: your full service list, your history, a slideshow, three competing buttons, or a video that autoplays.

Rewriting yours

A sequence that takes about half an hour:

  1. Write the answer to "what do you do and for whom" as one plain sentence, as if explaining it to a neighbour. Do not try to make it good.
  2. Cut it to twelve words. Cut adjectives first, then any word that would still be true if a competitor used it.
  3. Take everything you cut and see whether it belongs in the subhead. Most of it will not, and that is fine.
  4. Read the headline aloud. If you feel slightly embarrassed by how obvious it is, you are close.
  5. Run the five-second test on three strangers.
  6. If two of three get it, ship it.

The headline you end up with will probably feel too simple to you and completely unremarkable. That is what a headline that works feels like from the inside.

Frequently asked questions

Is twelve words a real rule or an arbitrary number?

It is a house heuristic, not a law — we use it because it is a limit you can actually check, and because headlines that run past it almost always turn out to contain two ideas that should have been a headline and a subhead. Usability research on how people read web pages consistently finds that they scan rather than read, and that the first few words of a line carry most of the weight. Twelve is our practical translation of that. Nine is better.

What if my business genuinely does several unrelated things?

Then the homepage headline names the category, and the page beneath it splits into services. "Commercial and residential electrical work in Arlington" is a fine headline for a business that does twenty different jobs. What does not work is trying to list all twenty above the fold. Each one earns its own page, which is also what Google needs.

Should the headline include my city for SEO reasons?

Include it because a human needs it, and take the search benefit as a side effect. A visitor who cannot tell in five seconds whether you serve their area has to work to find out, and some fraction of them will not. If the city genuinely does not fit inside twelve words, put it in the subhead directly underneath — it is still above the fold and still doing the job.

My headline is a tagline we have used for fifteen years. Does it have to go?

Not necessarily, but it probably has to move. Taglines are identity statements and they work on people who already know you — returning customers, referrals, anyone who arrived by typing your name. Strangers arriving from a search do not have the context that makes a tagline mean anything. The common fix is to keep the tagline near the logo and give the main headline to the plain-language version.

How do I run this test if I do not know anyone outside my industry?

Use someone in the building. The only real requirements are that the person does not work for you and has not seen the site. A neighbour, a supplier's front-desk person, a friend's teenager. Five seconds on a screen, then take it away and ask two questions: what does this business do, and would they serve you. Three strangers is enough to know whether you have a problem.

Does this apply to the pages that are not the homepage?

More than people expect. Most visitors from search land on an interior page, not the homepage, so every service and location page needs to pass the same test on its own. That is also why the H1 on each page should describe that page rather than repeating the business name — see the seven sections for how the rest of the page should follow it.

Sources