Most Small Business Sites Need Fewer Pages — and Clearer Jobs

Most small business websites need fewer pages than they are sold. The right count is not a magic number — it is an inventory of jobs a visitor can finish. Homepage decides, services prove the offer, contact converts, and every extra URL either earns a distinct job or becomes a distraction. This spoke sits under Websites That Feel Like Films.

The short answer

• Start with jobs, not a “recommended sitemap” from a theme demo or a 20-link competitor footer.
• Non-negotiable for most local and service businesses: Home, Services (or 2–4 service pages), About/proof, Contact.
• A service earns its own page when search intent, proof, and a distinct offer all exist — not because the brochure had a bullet.
• Bloated sitemaps dilute attention, slow launches, and leave half the pages half-written.
• Cutting pages is not cutting trust when the remaining pages carry real proof and a clear next step.

How many pages does a small business website actually need?

Enough to complete the path from “is this for me?” to “how do I hire you?” — and no more until a page has a job you can name in one sentence. Typical shipped counts below are outcomes of that inventory, not a target to hit.

Business type	Common outcome after a job inventory	What usually gets cut
Local trades (HVAC, plumbing, cleaning)	6–10 URLs	Generic “Our Process,” duplicate service blurbs, filler blog
Professional services (design, consulting)	5–8 URLs	Separate “Values,” “Team” snowflakes, unused case-study shells
Product / light ecom brand	6–12 URLs	Extra “Collections” with three SKUs, orphan campaign pages
One-location retail / studio	4–7 URLs	FAQ-as-page when a section would do, redundant location pages

If you cannot say what a visitor should do on a URL, it is not a page yet. It is a wish.

Agencies sell volume because volume looks like a proposal. Theme kits sell volume because every demo section wants a route. Neither is a brief. Nielsen Norman Group’s homepage research still holds: give people clear starting points for a small set of high-priority tasks — they name 1–4, not a dozen equals in the nav.

What is a job-to-be-done on a website page?

A page job is the change you want in the visitor before they leave: believe you, pick an offer, call, or book. It is not “have an About.” It is “trust the people enough to dial.”

Write the job as a sentence a stranger can repeat:

“A visitor on this page is trying to _____, and success looks like _____.”

Page	Job (the trying)	Success (the finish)
Home	Decide if you are the right contractor or studio	Tap Call, open Contact, or pick a service
Service	Confirm you fix their problem in their area	Call or request a quote
About / proof	Believe real people did this work	Click through to Contact or a service
Contact	Hire you with minimal friction	Form submit or phone dial
Service area	Confirm you actually go there	Call with the right expectation
Project / case	See yourself in a finished job	Click to the matching service or Contact

Jobs that fail the sentence test:

• “Tell our story” with no proof and no CTA
• “List everything we might do someday”
• “Rank for every city in the metro”
• “Look complete next to a competitor’s footer”

GOV.UK’s service manual starts forms with one thing per page so people can focus, recover from errors, and finish on a phone. Marketing sites are not government forms, but the same pressure applies: one finishable job per URL beats a binder dumped into routes.

Which pages are non-negotiable?

For most small businesses selling a service, the floor looks like this:

1. Home — who you are for, what you do, primary CTA (call or form).
2. Services index or 2–4 service pages — scoped offers with proof, not a menu dump.
3. About / proof — people, credentials, photos of real work (section or page).
4. Contact — phone, form, service area, hours, what happens next.

Everything else is earned. Gallery, FAQ, financing, careers, blog, city pages — useful when maintained; dead weight when empty.

URL	Why it earns a place	When it can stay a section
Home	First yes/no and the primary CTA	Never — you need one official home
Services	Buyer needs a scoped offer	One index if offers share process and proof
About	People and licenses reduce risk	Home block if the company is one person
Contact	Conversion and logistics	Almost never — hide this and you hide the phone
Work / gallery	Visual proof for trades and studios	Home or service page if you have six photos
Service area	Honest coverage for a traveling crew	Contact footer if you serve one town

NN/g’s homepage principles treat Home as the place that states purpose and prompts action. It is not a second brochure. If Home cannot decide, no amount of extra URLs will save the visit. Users often leave in 10–20 seconds unless the value proposition is obvious — a twelve-item nav is how you waste that window.

How do you inventory pages by job instead of by brochure?

Do this before anyone opens Webflow, Framer, or a theme. One hour. A whiteboard is enough.

1. List the offers that made money in the last twelve months — not the ones on the old PDF.
2. List the questions people ask on the phone before they hire you.
3. List the proof you actually have: photos, licenses, reviews that name a service, named people.
4. Write one job sentence per proposed URL.
5. Merge any two sentences that share the same trying and the same success.
6. Park the rest as a section, a download, or a later phase.

Input	Keep as a URL when…	Demote when…
Brochure section	It maps to a paid offer with unique proof	It is history, values, or “our process” with no CTA
Phone question	The answer needs room, photos, or a distinct CTA	A paragraph on Home or Contact would finish it
Manufacturer SKU list	You sell and staff that system as its own job	It is a spec sheet nobody searches
City list	You have unique proof and a maintainer	The paragraph is identical except the city name
Old blog	Someone owns a cadence and the posts answer buyers	The last post is fourteen months stale

Checklist before you approve the sitemap:

• Every URL has a job sentence
• No two URLs share the same success metric
• Every URL has a content owner and a proof source
• Nav labels match jobs, not internal department names
• Contact and the primary phone path survive any cut

If a proposed page fails two of those, it is not in v1. I have shipped hundreds of production sites. The ones that convert are the ones whose sitemap came from jobs, not from a binder.

Run the inventory as a one-hour session before design. Bring the person who answers the phone.

Minute	Output	Done when
0–15	Money offers	Three to five offers that actually billed
15–30	Phone questions	A list that matches last month’s calls, not the PDF
30–45	Job sentences	One sentence per proposed URL, no leftovers
45–60	Merge / wait / kill	A sitemap you can read out loud in sixty seconds

If the hour produces sixteen URLs, you did not inventory. You transcribed the brochure. Do the merge pass again.

What is the one-job-per-page test?

Force every proposed page through the job sentence. Then run the merge test: if two pages share a job, they are one page with sections.

Examples that pass:

Page	Job	Success
Home	Decide if you are the right contractor	Tap Call or open Contact
AC Repair	Confirm you fix their problem in their area	Call or request quote
Contact	Hire you with minimal friction	Form submit or phone dial
Project / case study	Believe you have done this before	Click to Contact or related service

Examples that fail:

• “Our Culture” with no hiring CTA and no trust payoff for buyers
• “Resources” with three undated PDFs
• “Services” that lists twelve items with the same paragraph rewritten
• “Commercial” and “Residential” that repeat the same three systems with a noun swap

Decision list:

• One job, unique proof — build the URL
• One job, shared proof — section on the parent page
• Three jobs on one URL — split or cut until one remains
• No job you can say out loud — do not build it

If you need a CMS collection to “make it easy to add pages later,” you are planning inflation. Webflow and every other CMS will happily mint URLs. That is not a reason to invent them.

When does a service deserve its own page?

Use this checklist. A service earns a URL when at least three boxes are true:

• People search for that service by name, or ask for it on the phone weekly
• The offer, price band, timeline, or process is meaningfully different from sibling services
• You have proof unique to that service (photos, reviews mentioning it, certifications)
• You would run ads or Google Business Profile categories aimed specifically at it
• A single combined Services page would bury it below the fold for the people who need it

Pass example (HVAC-shaped): AC repair, furnace install, and maintenance plans often deserve separate pages because intent, seasonality, and proof differ. On AllCity HVAC-style builds, the nav stays short — Services, Systems, About, Contact — and per-system pages exist because each system is a job: emergency repair is not a replacement quote, and both go cold inside an hour. That is a job inventory, not a manufacturer PDF turned into routes.

Fail example: “Duct cleaning,” “duct sealing,” and “duct inspection” as three thin pages with identical stock photos. One “Duct services” page with clear subsections usually converts harder and stays maintainable.

Signal	Own page	Stay a section
Search / phone demand	Named weekly	Mentioned once a quarter
Offer difference	Different crew, price, or timeline	Same visit, same quote shape
Proof	Photos or reviews that name it	Shared truck shots
Ads / GBP	You would spend against this term	You would not
Fold risk	Buried on a combined page	Visible in the first screen of Services

Thin siblings are not “more SEO.” Google’s people-first guidance asks whether someone who landed directly would leave having achieved a goal. Three 150-word clones fail that test on arrival.

Why do bloated sitemaps hurt conversion?

More pages feel like more professionalism. In practice they create unfinished surfaces and a nav of equals.

Failure	What it costs
Half-written pages	Visitors bounce on thin content; you look unfinished
Split attention	Homepage CTAs compete with twelve nav items of equal weight
Launch delay	Content becomes the critical path; the site sits in limbo
SEO theater	Thin service-area clones that Google ignores or treats as doorway junk
Maintenance debt	Nobody updates the tenth page; stale hours and dead CTAs linger

Google’s spam policies call out doorway abuse: pages built to rank for similar queries that funnel people to one real destination, or substantially similar pages that are closer to a search-results list than a browseable site. A city page that is the same paragraph with the city name swapped is that pattern. You do not need a manual action to lose. You already lost the visitor.

A short site with sharp jobs beats a long site with soft ones. Craft belongs in composition and proof — see the pillar — not in inventing URLs to impress a competitor’s sitemap.

How do you cut pages without cutting trust?

Trust comes from specificity, not page count. Keep trust; cut redundancy.

1. Merge siblings — one Services page with strong sections beats five 150-word pages.
2. Demote to section — testimonials, FAQs, and financing often live better on Home or Contact than as lonely URLs.
3. Archive honestly — unpublished drafts do not need public routes “for later.”
4. Keep one deep proof — a real project story outperforms three vague “portfolio” tiles.
5. Protect the phone path — never cut Contact clarity to save a line in the nav.

Cut	Where the job moves	What you must not lose
Team + History	About	Names, licenses, real faces
Financing	Contact or Home	Terms a buyer needs before they call
Thin service clones	Parent service page	The distinct offer, if it actually exists
Empty Blog	Remove from nav	Nothing — an empty Blog is a neglect signal
Extra city pages	One service-area page	Honest coverage and response expectations

Checklist before deleting a live URL:

• Nothing ranks that you care about, or you have a redirect plan
• The job moves to a surviving page with equal or better prominence
• Nav and footer no longer promise the old page
• Forms and call tracking still make sense on the new path
• You will ship a launch checklist that includes the redirect map, not a hope

Google treats a server-side 301 or 308 as a permanent move. Soft 404s and “we redesigned, good luck” are how you donate the few links you had. Point the old job at the page that now does that job — not at Home by default unless Home truly absorbed it.

What does a contractor sitemap look like before and after?

Sold as “complete” (16 labels): Home, About, Team, History, AC Repair, Heating, Installation, Maintenance, Indoor Air, Commercial, Residential, Gallery, Blog, Contact, Careers, Financing.

Ship-ready after a job inventory (8 pages):

Page	Job
Home	Qualify + call
AC Repair	High-intent emergency / repair
Heating & Install	High-intent replacement
Maintenance Plans	Retention / recurring
About	People + licenses + real trucks and jobs
Work / Gallery	Visual proof
Service Area	Where you actually go (one honest page)
Contact	Convert

Team and History fold into About. Financing becomes a Contact or Home section. Blog waits until someone will write monthly. Commercial stays off until you have commercial proof and a sales path. Indoor Air becomes a section on Maintenance or a later page when you have photos and a priced offer.

That cut is not “less business.” It is less fiction.

If you are a one-truck shop, you may ship six. If you staff commercial and residential with different proof and different forms, you may ship ten. The number follows the jobs. Do not start from sixteen and negotiate down by vibe.

When do service-area pages help, and when are they spam?

One honest service-area page — cities you actually serve, response expectations, a map or list — is usually enough for a single-location trades company.

Google’s Business Profile guidelines want a precise address and/or service area that matches how you operate. Service-area businesses get one profile for the central office and a designated coverage area — not a virtual office per suburb. Service-area and hybrid listings exist so customers can see where you actually go. Your website should tell the same story.

Add city pages only when:

• You have unique proof or offers per city (crew based there, reviews naming the city, different permits)
• Someone will maintain unique copy — not spun paragraphs
• You are willing to keep NAP consistent with the Business Profile and any LocalBusiness markup

Pattern	Verdict
One service-area page that lists real towns and hours	Build
City page with a local crew, local reviews, local permit notes	Build if you will maintain it
Same paragraph, city name swapped, stock skyline	Do not build — doorway-shaped
Hidden-address SAB pretending each city is a storefront	Do not build — it fights the Profile
Twenty ZIP pages “for SEO” with no dispatcher in those ZIPs	Do not build

If every city page is the same paragraph with the city name swapped, skip them. Thin location pages are a credibility tax, not a strategy. Markup cannot rescue a page that is not true on the screen.

Is a one-page site enough — and do you need a blog?

A one-pager works when:

• One offer, one geography, one CTA
• Proof fits without a scavenger hunt
• You are early and would rather ship than architect

It fails when services compete for attention, when searchers need distinct landing pages, or when industry visitors need a scannable About separate from the pitch. Most growing small businesses outgrow the pure one-pager before they outgrow a short multi-page set.

Situation	One page	Short multi-page
Single offer, one town, one phone number	Yes	Optional
Two or more high-intent services with different proof	No	Yes
You buy ads against named services	No	Yes — match the ad to a job
You are still collecting photos and reviews	Yes, then expand	Do not invent empty URLs
Editors will not maintain a CMS	Yes	Only if pages stay few

If you are on the fence, ship a short multi-page set. Expanding later is easier than untangling a scroll novel.

A blog is a collection with a job, not a page-count inflate. Add it only if someone owns a cadence and the posts answer questions people already ask on the phone.

Content type	Job	When it earns a URL
FAQ block	Unblock the hire	Almost always a section
Seasonal note on Home	Timely reminder (filters, storms)	Section until it is a real program
Project write-up	Proof for a finished job	Page when the story is specific
Blog index	Recurring answers	Only with an owner and a calendar

An empty Blog in the nav signals neglect. That is worse than no Blog.

How does page count affect quote price?

Studios price systems and content load, not “$X per page” in a vacuum — but page count still moves the number because each URL needs design states, content, QA, and often CMS fields.

Sitemap choice	Budget effect
5–7 clear pages, content ready	Baseline sprint
12+ pages, thin or missing copy	Content and IA become the project
Many near-duplicate service pages	You pay twice: build + rewrite later
One-pager now, structured growth later	Cheaper launch if scope is honest
City-page factory with no unique proof	You pay for pages Google will treat as junk

Ask for a quote against jobs, not against a competitor’s 20-link footer. If a proposal celebrates page volume, ask which pages share a job. That question saves more money than a discount.

What belongs in the nav is a separate decision from what exists as a URL. Footer and in-page links can carry secondary jobs without giving them equal weight in the header.

Item	Header nav	Footer or in-page
Primary phone / Contact	Yes	Yes — repeat it
Highest-intent service	Yes if it is a weekly hire	Link from Home if it is seasonal
About	Yes for most service firms	Optional if Home already proves people
Gallery	Only if proof is the closer	Usually a Home or service block
Financing / warranties	Rarely	Contact or service page
Careers	Only when you are hiring this month	Footer, or cut
Privacy / legal	No	Footer

How do you test a sitemap before you build it?

You do not need a research budget. You need five people who match the buyer and a tree you can point at.

Card sorting shows how buyers group offers — useful when you are arguing about labels. Tree testing then checks whether those people can find “request a quote for a dead furnace” in the proposed nav without the visual design to help them. NN/g treats sort as discovery and tree test as evaluation. Do both if the sitemap is contested. Do the cheap version if it is not: write the jobs on cards, ask a dispatcher or office manager to group them, then ask a homeowner where they would click.

Test	What you learn	Cheap version
Open card sort	How buyers group services	Index cards, 5 customers
Tree test	Whether labels find the job	A nested list in a doc, 5 tasks
Phone-script review	What people actually ask	Last 20 call notes
Proof audit	Which pages can be true	Folder of real photos and reviews

Five tasks that should succeed in a tree test for a trades site:

1. Find how to get a same-day repair.
2. Find whether you serve their town.
3. Find a phone number and what happens after they call.
4. Find proof you have done their kind of job.
5. Find the difference between repair and replacement.

If task 1 and task 3 fail, you do not have a sitemap problem. You have a business-model problem showing up as IA.

What breaks when the brochure becomes the sitemap?

A common failure: the owner hands over a 24-page PDF and asks for “a page for each section.” The site launches late, half the pages say the same thing, and the phone number is still in the footer in 11pt gray.

What it costs:

• Weeks of copy that nobody will maintain
• A nav that treats Careers and AC Repair as equals
• Thin URLs that sit in Search Console as discovered-not-indexed
• A launch that waits on the tenth page instead of the phone path

What to do instead:

1. Extract the three offers that make money.
2. Map one page (or one Home section) per offer.
3. Put credentials and people on About.
4. Put every conversion action on Contact and in the header.
5. Park the rest of the PDF as downloadable one-sheets if industry buyers need them — not as orphan URLs.

Brochure leftover	Better home
Mission / values	Two lines on About, or cut
1998–present timeline	One paragraph if a buyer cares; else cut
Staff bios with no photos	About, or wait for photos
Manufacturer spec sheets	PDF link from the matching service
“Why choose us” clichés	Replace with licenses, response time, real jobs

Bravery is shipping the short site that sells, not the long site that matches the binder. Rehearse the sitemap like a fold: one job per URL, then ship. Explore that model on /websites if you want the inventory built as a sprint.

FAQ

Do I need a blog?

No, not by default. Add a blog only when someone owns a publishing cadence and the posts answer real buyer questions. An empty Blog link hurts more than it helps. FAQ sections and project write-ups usually finish the same jobs with less neglect risk.

Should every service get a page?

No. A service gets a page when intent, proof, and offer differ enough that a combined page would bury the buyer. Thin duplicate service pages waste budget and attention. Use the three-of-five checklist before you mint another URL.

Do I need a testimonials page or a section?

Usually a section. Put the strongest quotes near the CTA on Home and service pages. A standalone testimonials page helps only when you have dozens of reviews worth browsing — and even then, the quotes still need to appear next to the hire action.

What about service-area pages?

One honest service-area page is enough for most single-location businesses. Add city pages only with unique proof and a maintainer — not spun copy for every ZIP code. Keep the page aligned with the Google Business Profile so you are not telling Maps one story and the site another.

Is a one-page site enough?

It can be for a single offer and geography. Most growing service businesses do better with a short multi-page set so services and contact do not compete inside one endless scroll. Ship the short set when two high-intent services already have different proof.

How does page count affect quote price?

More pages mean more content, design states, QA, and CMS work. Clear short sites are usually cheaper and faster than bloated sitemaps with half-finished URLs. Quote against jobs. If a proposal sells volume, ask which pages share a job.

CTA

Fewer pages. Clearer jobs. A site that answers the phone.

Explore /websites or book a Website sprint at /contact?intent=websites-sprint.