In brief
- Rhythm beats volume. One article a month for twelve months builds more authority than twelve articles published in six weeks and then nothing.
- Nobody writes alone. The in-house expert provides the material, the writer shapes it, AI speeds up the low-level tasks. In that order, never the reverse.
- The brief decides the final quality. A text comes back usable first time if it contains the search intent, the required outline and three in-house facts.
- Content is judged at six months. Before that, there is not enough data to decide anything.
In January, a small business decides to publish one article a week. The first quarter more or less holds. In April, the marketing manager has a trade show followed by two weeks’ holiday, and the blog stops.
It does not restart. Nobody decided to stop: the pace had simply been calibrated on January’s enthusiasm instead of March’s workload.
What decides whether an editorial strategy succeeds is rarely the choice of topics, but content production seen as an organisation: who writes, when, with what budget of hours. The strategic side, choosing topics and getting cited by engines, is covered in our article on combining SEO and content marketing.
Calibrate the calendar on your worst month, not your best
The right pace is the one you would keep up during the busiest month of your year. For most of the small businesses we support, that means one in-depth article a month. Give it a recurring date, the third Tuesday for example: a floating date becomes a “when I have time” that never comes.
Twelve articles published regularly carry more weight than forty concentrated in one quarter and then abandoned. A blog dormant for eight months stops being crawled frequently, and the team loses the habit: each restart costs more than the last.
Only increase the pace after two quarters held without exception. It is the only reliable indicator that the organisation can absorb the rhythm, more useful than any projection of your SEO, search ads and social visibility strategy.
Who really writes: three resources, none sufficient alone
The in-house expert has the material, not the time
Your technical manager, your managing director or your senior salesperson hold what makes content useful: the real figures, the objections heard, the cases that went wrong. They do not have three hours to write.
The rule that unblocks the situation: they do not write, they talk. Thirty minutes of recorded interview produce enough material for an in-depth article. An expert will give thirty minutes. Three hours of writing, no.
The external writer has the time, not the material
A professional writer structures, prioritises and makes things readable. Delivered with no raw material, they produce a correct, interchangeable text, the one your competitors will publish too. Their output depends on what you give them as input.
AI produces fast, and produces generic
What it does well: transcribing and summarising your expert’s interview, suggesting an outline, rephrasing a passage that is too technical, turning an article into five social posts. It cuts the time by two or three.
What it does badly: carrying a point of view, producing reliable figures, knowing your market. A fully generated article looks like the one from the competitor using the same tool, as we explain in our piece on realistic use of artificial intelligence in small businesses.
The combination that avoids mush: human material in, AI in the middle for speed, business review at the end. Remove the first or the last step and you get filler.
The writing brief, the document that decides everything
A text comes back usable first time, or after two rounds of corrections. Everything is decided before writing, in a one-page document.
- The reader’s exact question, worded as they ask or type it, not as a marketing title.
- What they should be able to do by the end of their reading: decide, calculate, prepare a meeting, rule out an option.
- Three in-house facts: an internal figure, an anonymised customer case, an objection actually heard in a meeting. That is what makes the text impossible to copy.
- The required outline as headings, plus the angle the text must not take, to avoid duplicating content already published.
- The internal links to place and the action expected at the end of the text.
- The format constraints: length, level of technicality, banned vocabulary.
What to remember Writing this brief takes thirty to forty-five minutes, against three times as long to rescue a badly framed text. Points 3 and 4 do most of the work: without in-house facts or a banned angle, the result will be average whoever writes it.
What each format really costs in time
A piece of content’s budget is counted in internal hours spent, approval included, and above all in how long the format keeps working.
| Format | Total effort | Useful life | Condition for keeping it up |
|---|---|---|---|
| In-depth article (1,200 to 1,800 words) | 6 to 10 h | 2 to 4 years if updated | Written brief, 30 min of an expert’s time |
| Customer case study | 8 to 12 h | 3 years and more | Customer’s agreement, before and after figures |
| Method page or offer page | 10 to 15 h | As long as the offer exists | Management decision |
| Practical guide or checklist | 4 to 6 h | 1 to 2 years | Careful layout |
| Newsletter | 2 to 3 h | 72 hours | A source piece already written |
| Social post | 20 to 40 min | 48 hours | An extract from an in-depth piece |
Read this table by the right-hand column. Hours invested in a format that lives 48 hours do not add up; those invested in an in-depth article keep producing for years.
Recycling: one in-depth article feeds six weeks
The rule fits in one sentence: write the longest format first. Starting from a social post forces you to rewrite everything to move up a format; the reverse only needs cutting up.
On publication day, prepare the newsletter that picks up the angle and links to the text, without copying it. It is the most profitable use of an email marketing strategy when nobody has time to write original newsletters.
In the following weeks, take three extracts from the same text: a figure, an objection dealt with, the checklist. Each becomes a post, with a different way in and a link to the article.
After a month, the questions asked in meetings on the subject become an extra paragraph in the original article, which grows thicker instead of ageing.
When to judge content, and on what
The most common mistake is to look at sessions after seven days and conclude that “the blog doesn’t work”. At seven days, an article has no usable data, whatever its quality.
| Milestone | What to look at | Possible decision |
|---|---|---|
| D+7 | Internal and sales feedback, factual errors | Correct, nothing else |
| D+30 | Impressions and entry queries in Search Console | Rewrite the title if click-through rate is low |
| D+90 | Stabilised rankings, reading time, assisted conversions | Enrich, merge with a duplicate, or let it live |
| D+180 | Recurring traffic, inbound links, external citations | Schedule an update or archive |
Also stop tracking page views alone. Two indicators tell the truth: the queries through which people arrive, and assisted conversions, the contacts who read an article before writing to you. The second requires a correct set-up of Google Analytics 4.
The mistake not to make Publishing without ever going back. A 2023 article stuck on page two often climbs back after three hours of updating, when the next one will take eight hours to start from scratch: history, inbound links and indexing already exist, only freshness is missing. Out of twelve slots a year, set three aside for existing content and start by auditing the content already online.
Your first editorial quarter, week by week
- Week 1: list the ten questions your customers actually ask in meetings. That is your starting calendar, before any keyword tool.
- Week 1: set the recurring publication date and block in the diary the expert’s two hours for the following month.
- Week 2: write the brief for the first topic using the six points above, then have it reread by a salesperson.
- Week 3: produce. Thirty-minute recorded interview, transcription, shaping, business review.
- Week 4: publish, then on the same day schedule the newsletter and the three posts drawn from the text.
- Week 5: write nothing new. Take a piece published a year ago and update it.
This cycle repeats: ten articles a year and three updates.
Frequently asked questions
How many articles do you need to publish a month to be visible?
One in-depth article a month, kept up for twelve months, is enough for most small businesses to rank on specific queries. What counts is regularity over a year, not monthly volume. Aim for a pace that seems too slow in January and that you will still keep in October.
Should you write in-house or use an external writer?
Both, with a clear split: in-house provides the material in a thirty-minute interview, the external writer writes and structures. A writer alone produces an interchangeable text; an expert alone never publishes, for lack of time.
Can you write your blog articles with generative AI?
For transcription, outlines, rephrasing and turning articles into posts, yes, the time saved is real. For substance, no: the tool knows neither your figures nor your customers. Keep a human at the start for the material and a human at the end for the business review.
How long before an article brings in traffic?
Allow thirty days for the first impressions, ninety days for stabilised rankings, six months to judge. That is why so many blogs are abandoned in the third month, just before the results arrive.
Is it better to write a new article or update an old one?
If an existing article already receives impressions without ranking well, updating it pays off faster than a new text, for a third of the effort: history, inbound links and indexing work in its favour. Set aside a quarter of your slots for this task.
The constraint to set now
Before choosing topics, set the production constraint: how many internal hours are available each month, and who gives them. Without that answer, the calendar stays theoretical.
The entry threshold is low: thirty minutes of an expert’s time, a one-page brief and a fixed date are enough to start, provided you treat those three elements as non-negotiable.
One last point, now decisive: your content is read by AI crawlers as much as by readers. Knowing which ones come to your site, and which ones you want to let in, is managed with AI Crawler Manager, published by Datafirefly Limited, our agency’s sister company. Producing content without knowing who reads it is writing into the void.
Unsure whether you can keep an editorial calendar going for twelve months? We start by counting the hours actually available in your team, then we identify the existing content to update before producing new pieces. Let’s talk.