Ask how much content a business needs and the usual answer is a cadence: two posts a week, one a week, four a month. It is a comfortable answer because it is measurable, and it is the wrong unit. Nobody has ever ranked because of a frequency.
The better question is how many things you need to be the best available answer to. That number is finite, it is smaller than the internet suggests, and for most local businesses you can write it down in an afternoon.
What content is actually for
Three jobs, and everything else is decoration.
- Be found for questions your customers ask before they are ready to buy. Someone searching how long a repair takes is not yet a lead, but they will be, and they will remember who answered.
- Answer the objections that stop people who are ready. Price, timing, what goes wrong, why you and not the cheaper option. These pages convert, and most businesses never write them.
- Give people something worth citing. Links, mentions and now AI citations go to the page that says something specific, not to the page that summarises what everyone already says.
Notice that none of these are served by publishing on a schedule. They are served by covering a defined set of subjects properly.
The set most local businesses need
Before a single blog post, the pages that carry the most weight are the ones businesses treat as admin.
- One page per service you actually sell, written as though it is the only thing you do — not a paragraph on a combined services page.
- One page per location you operate from, with the detail that makes it that location rather than any other.
- A pricing page, or an honest explanation of why the price depends. Refusing to discuss cost does not stop people needing the answer; it sends them to someone who will give it.
- Proof: case studies, before and after, real numbers where you have them and specifics where you do not.
- The questions you answer on the phone every week, written down once.
For a business with four services and two locations, that is fifteen to twenty pages. Done properly, it will outperform two years of weekly posts, and it is a finite project with an end.
A business with twelve pages worth reading beats one with two hundred worth skimming, and it is cheaper to maintain.
Then what?
Once the core set exists, publishing new pages is usually the second-best use of the effort. The first is improving what is already there — the page ranking on the second results page that needs one more section, the service page written three years ago that no longer describes what you sell, the article that gets traffic and has no way to contact you at the end of it.
Updating is unglamorous and it compounds. A page that already has some standing responds faster to improvement than a new page does to existing, which is the opposite of how most content plans are built.
What AI search changed, and what it did not
It did not change the need for content, and the widespread hope that it would has not survived contact with reality. What changed is which content earns anything. A page assembled from what other pages already say is exactly what a language model can produce instantly and for free, so it has no value to anyone, including you.
What retains value is the part a model cannot generate: what you have actually seen, what the job costs in your market, what customers here get wrong, the photograph of the real thing. That is also, conveniently, what persuades a human being. The same shift is covered in more detail in our writing on answer-engine optimization.
Who should write it
The specifics have to come from inside the business, because that is the only place they exist. What an agency or a writer contributes is structure, search demand, editing, and the fact that it actually gets finished — which is the step that fails most often.
A practical arrangement: someone senior talks for thirty minutes about a subject they know, and a writer turns it into a page. That produces something no competitor can copy, and it costs an hour a month rather than a standing obligation nobody meets. If you would rather hand the whole thing over, that is what a dedicated marketing expert does — and where you should start depends on which of the core pages you are missing, which a free assessment will tell you.
Frequently asked questions
- How often should a small business publish?
- As often as it has something specific to say and no more. A well-researched page each month beats four thin ones, and both beat a schedule abandoned in April. If you must have a cadence, make it monthly and spend the rest of the effort improving pages that already exist.
- How long should a blog post be?
- Long enough to answer the question completely and no longer. Length correlates with ranking mainly because thorough answers tend to be longer, not because word count is rewarded. Padding a six-hundred-word answer to two thousand makes it worse for the reader and no better for search.
- Is AI-written content bad for SEO?
- The tool matters less than whether the result says anything. Content that only restates what is already published competes with an infinite supply of the same thing, however it was produced. Used to draft and structure material that comes from your own knowledge, these tools are simply faster writing.
- When should we stop adding pages?
- When every service, location and common question has a page that genuinely answers it. At that point the return moves to depth — improving, updating and consolidating — and adding pages starts to dilute rather than build. Very few small businesses reach this point, which is itself informative.