[//] Conversion

Writing a website people actually read

3 minute read · Updated 1 August 2026

Nobody reads a website the way they read a book. They scan, in a rough F-shape, looking for the thing they came for. Anything that gets in the way of that is friction.

Good web writing accepts this rather than fighting it.

Put the answer first

Most business pages build up to the point. Web pages should open with it. The reader arrived with a question — answer it in the first line, then explain.

Compare "Established in 2019, we are a family-run business committed to excellence" with "We fix boilers across Leeds, usually same day." The second tells a stranger whether they are in the right place. The first tells them nothing they can act on.

Your history is not irrelevant, it is just not the opening. It belongs further down, where someone already interested goes looking for reassurance — see building trust.

Structure for scanning

Real headings, short paragraphs, and lists where a list is honest. Three lines is a long paragraph online. A wall of text gets skipped in its entirety, regardless of how good it is.

Use headings that describe the content underneath, not clever ones. Somebody skimming should be able to read only your headings and get the gist.

This structure does double duty: it is also what search engines use to understand the page, as covered in technical SEO basics.

Write how your customers speak

Every industry develops language that makes sense internally and means nothing outside. If your customers say "boiler broken" and your site says "domestic heating system diagnostics", you are invisible to the people searching.

The practical method is to use the words your customers use in emails and phone calls. Those are the same words they type into Google, which makes this both a clarity improvement and an SEO one.

Avoid the filler that says nothing: "solutions", "cutting-edge", "passionate about quality". If a sentence could appear on any business's website, it is doing no work on yours.

Give every page one job

A page trying to explain five services, tell your story and collect enquiries does none well. One page, one purpose, one obvious next step.

It is also better for search: a page about one thing can rank for that thing, whereas a page about five competes with itself. See navigation and structure for how to split things sensibly.

Length, and the myth about it

There is no magic word count. Long content is not automatically better; it correlates with ranking mainly because thorough answers tend to be longer, not because length itself is rewarded.

Google's own guidance on helpful content (opens in a new tab) is about whether a page satisfies the reader, not how many words it contains. Write until the question is answered, then stop.

Sources and further reading

Common questions

How long should a web page be?
Long enough to answer the question properly and no longer. A service page might be 300 words; a detailed guide might be 1,500. Padding a page to hit a word count makes it worse, not better.
Should I write the content or pay someone?
You know the work and the customers, which is the hard part. Many people write a rough version themselves and pay someone to tighten it. That is usually better value than briefing a writer from nothing.
How often should I update my content?
When it stops being accurate — prices, services, areas covered. Rewriting for the sake of freshness achieves little. Genuinely outdated information is worse than old-but-correct information.

Read next

Findability3 min

Technical SEO basics

How search engines find and understand your pages, and the technical faults that quietly keep you out of results.

Conversion3 min

Building trust

The signals that make a stranger believe you are real, competent and safe to contact.

Or see all 19 guides — or head back to the Smiths SEO home page to see what I do.

Not sure how your site measures up? Find out for free.

Send me your web address and I’ll go through it properly, then write back within 24 hours with what’s broken, what’s slow, and what it would cost to put right. Nothing to pay unless you decide to go ahead.