Content strategy in the local services market has been distorted by advice written for publishers. Post weekly. Build topical authority. Produce a content calendar for the year. For a media business with an audience to feed, all of that is sound. For a dental clinic, a renovation firm or a law practice in Valencia, it usually produces a blog nobody reads, attached to service pages that were never improved.
This article describes the approach we use instead: a finite set of pages that cover every way a customer might arrive, built once, maintained deliberately, and measured by whether they produce enquiries rather than by how many were published.
A local service business has a bounded market. There are only so many ways someone can search for a bathroom renovation, an employment lawyer or an orthodontist in a city. Once you have covered them, publishing more does not extend your reach — it starts competing with what you already have.
This is genuinely different from publishing, and the difference has budget consequences. A publisher needs a content engine. A local business needs a content project followed by a maintenance habit.
| Page type | How many | Purpose |
|---|---|---|
| Service pages | 4 - 10 | One per service someone searches for separately. The pages that convert. |
| Decision-support pages | 3 - 6 | Pricing, process, comparisons, what to look for. Where trust is won. |
| Proof pages | 2 - 5 | Cases, before and after, credentials, team. Rarely rank alone, decisive for conversion. |
| Question pages | 5 - 15 | The real questions customers ask before buying. Feed both search and AI answers. |
| Local pages | 0 - 6 | Only where you have genuinely district-specific material. |
Fifteen to twenty-five pages, done properly, cover the entire realistic search market of most local businesses in Valencia. That is a three to six month project at a sane pace, not an indefinite subscription to weekly blogging.
Keyword tools tell you what people typed. They do not tell you what people meant, what they were worried about, or what stopped them from buying. Three better sources, in order of value:
Your phone and inbox. The five questions every prospect asks before committing, in their exact words. This is the highest-value content input available to any local business, it costs nothing, and almost nobody harvests it. Ask whoever handles enquiries to write down every question for two weeks.
Your own impressions. If the site has been live more than six months, you already appear for queries nobody chose. In Search Console analytics, the queries with impressions and few clicks are confirmed demand you are failing to capture — and the vocabulary is real rather than modelled.
Conversational questions. The long, condition-laden questions people put to AI assistants are a genuinely new demand layer and they do not appear in classic keyword tools. Researching them in AI Analytics regularly surfaces commercial questions in local categories where nobody in the city answers convincingly.
A notebook by the phone for a fortnight produces a better content plan than a month of keyword research. Every recurring question becomes either a section on a service page or a page of its own, written in the customer's own words rather than in industry vocabulary.
The most common way local content programmes fail is by producing several pages for the same intent. "Bathroom renovation Valencia", "bathroom refurbishment Valencia" and "bathroom remodelling company Valencia" are the same question in different words. Three pages compete with each other, split internal links, and each performs worse than one good page would have.
The test is not whether the wording differs. It is whether a customer asking those two things wants a different answer. If the answer would be the same page, it is one page.
The exception is when the search results themselves show two intents coexisting — if the top ten mixes guides and service pages, both kinds of page have a place, and they should link to each other clearly.
Length is not the variable. Specificity is. A twelve-hundred-word page with no concrete information loses to a six-hundred-word page that answers the question. Five elements make the difference:
The question resolved in the first paragraph, without preamble. Both readers and AI systems reward it.
Prices or ranges, timeframes, conditions, coverage. The things competitors leave out.
Headings phrased the way people ask, short blocks, extractable answers.
Real photographs, named cases, numbers from your own experience rather than industry averages.
The fifth is a review date. Content with a visible last-reviewed date is treated more seriously by readers, and it forces the maintenance habit that most content programmes lack entirely.
This is the least glamorous section and the one with the highest return. In most local sites we audit, the biggest available gain is not a new page — it is the ten existing pages that have decayed.
Decay has three causes. Prices and processes changed and nobody updated the page. Competitors published something better and the comparison shifted. Or the search results themselves changed shape — a new block appeared, and the same position now yields half the clicks.
The maintenance routine is simple and quarterly. Compare two 28-day periods at URL level, sort by lost clicks, and take the ten biggest losers. For each one, read it as a customer would and ask what is now out of date, missing or worse than the alternatives. Rewriting one decayed page that already has authority reliably outperforms publishing two new ones.
The content question businesses argue about is what to write. The question that actually determines the outcome is who writes it, because the difference between a page that ranks and one that does not is usually the presence of information only an insider has.
Three models, with honest trade-offs. Written internally by the person who does the work produces the most differentiated content and the least reliable schedule. The plumber who writes the page knows why a 1960s building in the old town needs a different approach - and will not write anything in the three weeks when work is busy. Written externally by a specialist produces a reliable schedule and generic content, unless the specialist is given the raw material to work from. Interviewed and written up is the model that works: someone external spends forty minutes recording the expert, then writes. It costs more per page than pure outsourcing and produces something competitors cannot copy from the same public sources.
That interview model has an additional benefit that is easy to overlook. It captures the vocabulary. Experts describe their work in words that are almost never the words in a keyword tool, and those phrasings are frequently exactly what customers type - because customers heard them from a previous supplier.
Whichever model you use, one rule holds: someone with actual domain knowledge must approve every page before it publishes. In regulated sectors - health, legal, financial - this is not an editorial preference, it is a liability question. Text that is merely plausible is a risk in those fields, and it is precisely what unreviewed generated content produces.
A page with no internal links pointing to it tells search engines that you do not consider it important. It is the most common reason a genuinely good page underperforms, and it is entirely free to fix.
Three rules cover most situations. Every new page should receive at least two links from existing pages that already have traffic, added at publication rather than "later". Anchor text should describe the destination rather than say "click here" or the bare brand name. And question pages should link to the service page that resolves the question, using the language a buyer would use.
The reverse direction matters too: service pages that link out to the question pages supporting them build a visible cluster rather than a set of isolated documents. This is what people mean by topical authority in a local context, and it requires no external links at all.
In Valencia this question has a practical edge. A page written for a local homeowner and a page written for an international buyer of the same service are not translations of each other — they answer different worries. The local reader wants price, timing and references. The international reader wants to know whether you work in their language, how the process differs from their country, and what happens if they are not physically here.
Before committing to translated content, check whether the demand exists: the country and language breakdown of your existing impressions is first-party evidence and takes a minute to read. In tourism, accommodation, dental, cosmetic and property-related sectors the answer is frequently yes and it goes unnoticed for years.
If the demand is there, localise the four or five commercial pages rather than translating the whole blog. A translated, unmaintained blog adds nothing and creates a maintenance burden nobody signed up for.
The wrong metric is pages published. The right sequence is four checks, each of which isolates a different failure:
Four questions, four different actions. Without them, "the content is not working" is an unanswerable complaint. With them, it is a diagnosis — and all four data points sit in the free layer of Semalt, sharing one period and one domain.
Your impressions with few clicks are confirmed demand you are not capturing. It is the cheapest content plan available, and it is already in your own data.
Sign in to Semalt See AI AnalyticsTwo to four good pages a month during the build-out phase, then maintenance. Weekly publishing forever is advice written for publishers, not for service businesses.
You need pages that answer buying questions. Whether they live under /blog/ is a URL decision, not a strategy. Many local sites would do better with more service and question pages and no blog.
No. Specific is better. A short page that states prices, conditions and timeframes beats a long one that says nothing checkable.
As a drafting aid with your own facts, prices and cases supplied, yes. Published unedited, it produces exactly the interchangeable text that cannot rank, because it contains nothing only you could have written.
When every buying intent has a page and the maintenance routine is running. After that the work is updating, not adding.
Content strategy for a local business is a bounded project with an open-ended maintenance tail, not an endless publishing commitment. Cover the intents, write pages that contain things competitors leave out, link them properly, and then spend most of the ongoing budget keeping the best ones current.
The cheapest starting point costs nothing and takes an hour: open your queries with impressions and few clicks over the last ninety days, and pick the ones with clear commercial intent. Connect the domain and pull that list — in most businesses we work with, it produces the first five pages of the plan without any further research.
If you want the full page map built and prioritised, including the cannibalisation check against your existing pages, get in touch. It is part of the free initial audit we deliver within 48 hours.
Free analytics for Google Search, SERP rankings and AI answers - plus fast indexing for your new pages.
Sign in to Semalt semalt.comGet in touch for a free audit
