Answer in brief
Invented topics do not coincide with what people search for. Five sources of topics inside the work you already do, a realistic publishing frequency, and three reasons plans collapse.
A site ages from silence, not from age
A site where nothing has changed for a year looks abandoned regardless of how well it was built. It reads in the small things: a latest news item from two years ago, a promotion that ended, the year in the footer.
Worse: search engines read it too. A section nothing is added to gets crawled less often, and new pages inside it index more slowly.
A content plan is not there so you can run a blog. It is there so the site answers the questions you get asked, and so those answers are findable in search.
Below: where to find topics you do not have to invent, how much to actually publish, and why plans collapse in the third month.
What a content plan is and what it is not
A content plan is a list of what will be published, with dates and owners. That is all.
It is not a list of topics. A topic without a date and a person is an idea, and everybody has plenty of ideas; they do not turn into text on their own.
It is not about a blog. What should be updated first is the commercial pages: service descriptions, prices, answers, case studies. Articles are one format among several and not the fastest to return anything.
And it is not about quantity. A plan for a hundred publications of which eight get made is worse than a plan for twelve that all appear.
A practical definition: a plan is what can be delivered with current resources, not what you would like.
It starts with questions, not topics
The commonest mistake is sitting down to invent topics. Invented topics do not coincide with what people search for.
Start with the questions you already get. Collect a month of email and calls and write down everything you were asked. Twenty questions accumulate in an hour.
That list immediately shows two kinds of material: what people ask before buying (what does it cost, how long does it take, what is included) and what they ask afterwards (how do I use this, what if).
The first answers for enquiries, the second for people coming back and recommending you. Start with the first.
And check against your own analytics: which queries already find you and which pages people leave without acting. That is a second source of topics, and it is more honest than any brainstorm.
Where topics come from without inventing them
Five sources, all inside the work you already do.
Client questions before a deal. The most valuable source, because behind each question stands somebody prepared to pay.
Objections. What stops people buying: too expensive, too slow, unclear, not sure. Every objection is material that removes it.
Mistakes you unpick for clients. Everything you regularly fix on other people's behalf is a ready topic, and it demonstrates competence immediately.
Your own data. Figures from your own work convince more than anybody else's. Our indexing analysis was built on our own report: 738 pages indexed out of 3310. That cannot be invented or copied.
And queries where you already appear but get no clicks. That is a ready list of where you almost reached.
How much to publish
A question people answer with a frequency when it should be answered with a resource.
Regularity beats volume. One piece a fortnight sustained for a year works better than eight in a month and silence afterwards.
A realistic minimum for a small company is two pieces a month. That is twenty-four a year, and it is already a meaningful quantity.
There will be no return in the first months. The search engine has to find the pages, assess them and decide they beat what sits above. Six months is a normal horizon to first results.
And do not chase word count. You need enough to close the question and not one line more. Length on its own does not help — completeness of the answer does.
Who writes
The place where plans break most often.
Option one: the owner or a specialist writes. The texts come out substantial and appear irregularly, because that person has a day job.
Option two: a hired writer. Regular and without expertise, unless you supply the material. A writer working without input from you produces a retelling of somebody else's articles — exactly what the search engine has already seen.
The arrangement that works is hybrid: the specialist supplies the substance by voice, the writer turns it into text, the specialist checks the facts. Half an hour of conversation instead of three hours of writing.
And appoint one person responsible for publishing. Not for writing — for the piece actually appearing on time. Without that role a plan becomes a wish list.
What to publish besides articles
Formats that return something faster than long texts and demand less.
Answers to questions on the service pages themselves. Five real questions with short answers do more for conversion than a two-thousand-word article.
Case studies. What it was, what you did, how it turned out, how long it took. The format that convinces best and writes faster than an article.
Updates to prices and terms. Not content in the usual sense, but these are exactly the changes both visitors and search engines are waiting for.
Photographs of the process. They need no text at all and build trust better than descriptions.
And short notes on what changed: a new service, a new timescale, a new limitation. That is what keeps a site alive, rather than a long article once a quarter.
What to do with old material: Twenty client questions accumulate in an hour and give…
A section missing from almost every plan, and its return is higher than that of new writing.
A two-year-old piece that brings people in is cheaper to update than a new one is to write. It already has positions, links and accumulated trust.
Updating is not changing the date. It is new figures, current examples, removing statements that have gone stale, and adding answers to questions that have arisen since.
Once a quarter, open the list of pieces bringing traffic and update the top three. That is a few hours of work with a predictable effect.
And look separately at pieces that have brought nothing for two years. Either rewrite them around a different question or remove them — weak pages damage the reputation of the whole section.
How to tell whether it works
Three measures. No more are needed.
The number of pages producing at least one search visit a month. Rising means the plan is working; flat means the topics are the wrong ones.
The number of enquiries arriving from pages published in the last six months. That is the direct link between the plan and the money.
And queries that started finding you and were not there before. Those show you have taken new territory rather than rewritten what you already had.
Look once a month and no more often. Console data lags several days, and the effect of a publication emerges over weeks.
Why plans collapse
Three reasons, and all three are predictable.
The first: the plan was made at peak enthusiasm. Twelve pieces a month look achievable in January and are already missed in February. Plan for your worst month, not your best.
The second: nobody owns publishing. When everybody is responsible for the deadline, nobody is, and the piece slips indefinitely.
The third: writing about the wrong thing. We saw this on our own journal: two articles about well-known people produced nine hundred and sixty-nine impressions out of a little over a thousand and exactly zero clicks. The audience arrived, but they had not come for services.
Hence a rule: if a piece does not answer a question people ask before buying, it can be interesting and still useless to the business.
A year's plan in one table
What the table needs to contain to work.
The publication date. Not a month but a specific day, otherwise everything drifts to the end of the month and does not appear.
The question the piece answers. A question, not a topic: it determines both the text and the heading.
The format: an answer on a service page, a case study, an article, an update to something old.
The person responsible for publishing and the person responsible for the substance — often two different people.
And a done column with a date. In six months it will show your real pace, and the next plan will be built on that rather than on the pace you hoped for.
Practical checklist
- Collect a month of client questions from email and calls.
- Split them into what is asked before buying and what after.
- Check which queries already show you without producing clicks.
- Set specific publication dates rather than months.
- Appoint a separate owner for publishing on time.
- Update the three highest-traffic pieces once a quarter.
- Add a done column with a date so you can see your real pace.
Questions and answers
Where should a content plan start?
With the questions you already get, not with invented topics. Collect a month of email and calls and write down everything you were asked: twenty questions accumulate in an hour. Invented topics do not coincide with what people actually search for.
How often should you publish?
Regularity beats volume. A realistic minimum for a small company is two pieces a month, twenty-four a year. One piece a fortnight sustained for a year works better than eight in a month followed by silence.
Who should write the content?
The arrangement that works is hybrid: the specialist supplies the substance by voice, a writer turns it into text, the specialist checks the facts. A writer without your input produces a retelling of other people's articles, and a specialist writing alone publishes irregularly.
Is it worth updating old material?
Yes, and the return is higher than from new writing. A two-year-old piece already has positions, links and trust. Updating means new figures and current examples, not a changed date. Once a quarter, update the three pieces bringing the most traffic.
Why do plans stop being followed after two or three months?
Three reasons: the plan was made at peak enthusiasm and not sized for the worst month; nobody owns publishing on time; and the writing is about the wrong thing — interesting, but not answering a question people ask before buying.