How Many Pages Does a Small Business Website Need? Count Questions, Not Sections

A short note on scope before the answer. What the site is for — the five jobs it gets hired to do — is worked through in what a small business website should actually do, and this post assumes you have read or at least skimmed it. How page count moves the price of a build sits in our small business website cost guide. Neither is repeated here.

The question this post answers is narrower and more practical: when you sit down with a blank sitemap, how many boxes should be on it, and how do you decide whether a given topic is a page or a paragraph?

Why “Five Pages” Is the Wrong Starting Point

Most owners arrive with a number already in mind, and it nearly always came from a template: Home, About, Services, Gallery, Contact. That structure is organised around what a business is, which is how the owner thinks. Buyers do not search for what a business is. They search for an answer to a question they have right now — who fixes a leaking flat roof near me, how much does a crown cost without insurance, can a moving company take a piano up three flights.

A page is a container for one answer. A five-page site built around company sections can be the best answer to almost nothing, because its single “Services” page tries to answer eleven different questions at once and is the strongest result for none of them. The page count that matters is the number of distinct questions your best customers ask before they buy, plus the handful of pages every business needs regardless.

The Split Test: When a Topic Deserves Its Own Page

The decision that actually sets your page count is made dozens of times during planning: is this one page or two? We use a three-part test, and a topic gets its own page only if it passes all three.

1. Is the searcher different?

Different words typed by a different person. A homeowner searching for emergency drain cleaning at 10 PM is not the property manager researching annual jetting contracts, even though both are “drains.”

2. Is the proof different?

The photos, examples, reviews and credentials that convince this buyer. If the evidence that persuades one audience would be irrelevant or off-putting to the other, the topics want separate homes.

3. Is the next step different?

Book a slot now, request a site visit, download a scope, call a specific person. When the action you want differs, a shared page forces one of the two audiences down the wrong path.

Three yeses make a page. One or two make a section. This single rule prevents both of the common failures: the cramped site where unrelated services share a page, and the bloated site where near-identical services each get a thin page of their own. Run it honestly and the page count stops being a preference and becomes a consequence.

The Floor: Pages Every Small Business Needs Anyway

Before the split test runs on your services, a small fixed set exists on nearly every site we plan, because each one does a job no other page can.

A homepage that routes visitors to the right answer within a few seconds rather than trying to be every answer. An about page with real people, real history and real reasons to trust you — it matters far more to referred customers than owners expect. A service-area page that states plainly where you work and where you do not. A proof page of specific results, completed jobs or case notes, rather than adjectives. A contact page that works after hours. And for most service businesses, a page that explains how pricing works, even if it publishes ranges rather than numbers.

That is six. Everything above it is driven by your services and your buyers, which is where the range of eight to twenty-five comes from.

Three Shapes of Business, Three Different Counts

One offer, one kind of buyer: eight to ten pages

A window-cleaning company, a single-discipline therapist, a specialist bakery. The floor plus two to four pages answering the questions that genuinely differ — residential versus commercial, a common objection, a seasonal service. Padding this kind of business out to thirty pages produces thirty versions of the same answer.

Several service lines: twelve to twenty-five pages

A plumbing and heating company, a dental practice, a general contractor. Each service line the business actually wants more of usually passes the split test, and the strongest ones often justify a supporting page for their most common question. This is where most of the small businesses we work with land.

Referral-led: one to five pages

An accountant with a waiting list, a consultant who only works through introductions. The site’s job is verification, not discovery, and one excellent page can do it. The mistake here is not too few pages but a thin one — a referred buyer checking you out wants enough substance to forward to a business partner with confidence.

Where Page Count Goes Wrong in the Other Direction

A large share of the bloated sites we audit were not bloated by the owner. They were bloated by a proposal that sold pages by the dozen, usually in one of two forms.

The location grid. A page for every suburb within thirty miles, each identical apart from the place name. Search engines recognise the pattern and treat the set as low value, and in our experience the grid rarely outranks one strong service-area page plus a few genuine location pages for the markets that matter. A location page earns its place only when it carries something true about that place.

The keyword split. Separate pages for “roof repair,” “roof repairs,” and “roofing repair company,” written to catch phrasing variations that search engines already understand as the same question. They compete with each other and dilute the one page that should have won.

Every page is a promise to keep it current. Opening hours, staff, prices and services change, and a page nobody maintains drifts out of date. That drift costs more now than it used to, because AI assistants read your site alongside your listings and lose confidence when the two disagree — the mechanism is laid out in what AI assistants actually cite. A twenty-page site you keep accurate beats a sixty-page site you cannot.

After Launch: Grow From a Question Log, Not a Plan

The launch set is a starting hypothesis. The pages worth adding afterwards announce themselves, provided you are listening in two places.

The first is your front desk. Keep a running list of the questions customers and prospects ask by phone and email. Any question that turns up every week and has no page is a page waiting to be written, in the words the customer used. The second is your search data. Once the site is live, search reports show queries you appear for without having a page that properly answers them. When a query keeps showing up and the matching page is a passing mention, that topic has passed the split test on evidence.

Review the list quarterly. Add the two or three strongest candidates, merge anything that has started to overlap, and remove pages for services you no longer want. That rhythm keeps a site growing in the direction customers are pulling it, which is the core of how we run SEO for small businesses and answer engine optimization.

Why a Quote Priced Per Page Tells You Little

Many proposals put page count at the centre of the price, and it is a weak unit. A page assembled from a template and a page written from an interview with your best technician are both “one page,” and the second can be worth more than ten of the first. When a vendor asks how many pages you want, the better reply is to hand over the list of questions you want the site to answer and ask them to propose the structure. Their answer tells you whether they are planning a site or filling a template. Why two proposals for the same business can differ by four times is unpacked in the vendor quote gap.

How We Plan a Sitemap

Every website we build starts with the question list rather than a page list. We collect what customers ask, run the split test on every candidate topic with the owner in the room, and only then draw the sitemap. Because we build with Claude Code, adding a page that passes the test later is quick and inexpensive, which removes the main reason businesses over-build at launch: the fear that every future page will mean another expensive project. You can launch lean and grow from evidence.

If you want one exercise to take away, write down the ten questions customers ask most often before buying, run the three-part split test on each, and count the yeses. Add six for the floor. That number will be closer to right than any template.