A Spring Hill Plumbing Site and the Case for Fewer Pages

We usually argue for more pages. Give every service its own page, write it properly, stop making one crowded services page carry six searches. That advice is right most of the time. This is the case where it was wrong, and thinking about plumbing website structure through the exception turned out to be more instructive than another example of the rule.

The site belonged to a plumbing company in the Spring Hill area, described anonymously here as we do with every teardown. It had 34 pages. It should have had about nine. Cutting it was the single most useful thing we did.

How a site ends up with 34 pages

Nobody plans this. It accumulates. Someone read that more pages help search visibility, which is true in the sense that more good pages help. So the site grew a page for every service, then a page for every service in every town, then a page for every combination.

Water heater repair. Water heater repair Spring Hill. Water heater repair Brooksville. Water heater installation. Water heater installation Spring Hill. And onward, through drains, repipes, and fixtures, until the navigation had four levels and the sitemap had 34 entries.

The pages averaged well under 200 words. Most were the same three paragraphs with the town name swapped, which is a pattern search engines have recognized for a very long time. Thin content is common in our audit corpus, where 7 of 45 sites had homepages under 200 visible words, but this site had reproduced that thinness thirty times over.

What that actually did to the site

Three things, all of them bad, and none obvious from the owner's chair.

First, the pages competed with each other. When five pages address roughly the same query, a search engine has to choose one, and it frequently chooses badly or splits the value between them. The company had built its own competition.

Second, the site became unmaintainable, which is the failure that lasts. Thirty-four pages means 34 phone numbers to update, 34 places for a stale claim to hide. This site had accumulated the usual evidence of that: a footer year three years behind, which puts it with the 23 of 45 sites in our corpus carrying a stale copyright, and leftover template copy on several interior pages, matching the 12 of 45 we find with visible placeholder text.

Third, navigation became a maze. Four levels of menu on a phone is not navigation, it is an obstacle course. The site also carried dead hash-only links, the kind that look tappable and do nothing, which we find on a smaller share of sites and which almost always live on the pages nobody has opened in years.

What we cut, and what we kept

The rule we applied was simple: a page survives if someone would search for it in words that differ meaningfully from another page's words, and if we could write several hundred honest words about it.

• One page per genuine service, written at real depth: drains, water heaters, repipes, leak detection, fixtures
• One emergency page, promoted to top-level navigation with a tappable number
• One service area page naming every town served, rather than a page per town
• A homepage that said what the business did and where, within the first screen

The town-by-town pages all collapsed into that single service area page. This is the part owners resist hardest, because those pages feel like coverage. In practice a set of near-identical pages differing only by place name is the pattern most likely to be treated as low-value, and one honest page naming your real service area does the job better.

The surviving service pages absorbed the content worth keeping and then got substantially expanded, because the material for five thin pages, combined and then properly written, makes one good one.

The fixes that mattered more than the cutting

It would be neat to say the restructuring did the work. It helped, but two smaller repairs mattered more immediately.

The phone number was plain text with no click-to-call link, putting this site with the 20 of 45 in our corpus that fail that test. It is the costliest defect we find and the one we fix first, and on a plumbing site, where a meaningful share of visits are urgent, it is the most expensive thing on the page. It took minutes to fix.

The site also had no viewport meta tag, one of 10 such sites in our corpus, so the whole four-level menu was being rendered as a shrunken desktop layout. Adding that single line improved the navigation more than the restructuring did, before we removed a single page.

When more pages really is the answer

We do not want this read as an argument for a small site. The advice that each real service deserves its own page is still correct, and most contractor sites we see have too few pages rather than too many.

The distinction is whether a page exists because a customer searches for that thing in their own words, or because someone was generating pages to have more of them. A page for water heater installation is real: people search exactly that. A page for water heater installation in each of nine towns is not, because the customer's words did not change, only the place name did.

The other test is maintenance. If you cannot keep a page accurate, you should not have it. Every page is a small ongoing obligation, and a site that has outgrown its owner's capacity to maintain it will go stale everywhere at once. Nine pages kept current beat 34 abandoned ones, every time.