Publishing content regularly without spending hours on it
Almost every small business I know has the same story with social media and the blog: it starts with enthusiasm, publishes three weeks running and, when the first busy period hits, the account goes silent for months. And an abandoned profile makes a worse impression than none: whoever looks you up wonders whether you are still open.
The problem is not a lack of ideas. It is that publishing well takes time the business does not have: thinking of the topic, writing it, adapting it to each channel, finding the moment. If that depends on "someone having a spare moment", it will never happen. The good news is that regularity is not achieved with willpower, but with a system.
Regularity beats perfection
Before getting into how, an important change of mindset: for a shop, a workshop or an advisory firm, publishing one useful piece every week is worth more than a brilliant campaign twice a year. Search engines reward live websites, and customers trust a business that shows signs of life. You do not have to be witty every Tuesday; you have to be there every Tuesday.
A system in three pieces
1. A list of topics that never runs out. The content that works best for a small business is not industry news: it is the questions customers already ask you at the counter or on the phone. "How often does this need servicing?", "what does this service include?", "what is the difference between X and Y?". Note them down for a week and you will have topics for six months. Every repeated question is an article, a post or a short video.
2. AI drafts; you supply the judgement. This is where the hours come back. A well-built content system takes your list of topics, your house tone and your examples, and prepares the drafts: the blog article and the short versions for social media or the newsletter, each adapted to its channel. The work left for you is the work nobody can do for you: reading it, correcting what does not sound like you, and deciding whether it gets published. Reviewing a text is ten minutes; writing it from scratch, two hours.
3. A calendar that runs itself. Once the rhythm is decided — one article a week, two social posts — the system presents the proposals with lead time, you approve them and they are scheduled. No more "I'll do it on Sunday".
The example: a homeware shop
A homeware shop had everything it takes to make good content: the owner can answer any question about choosing a pan or looking after a knife, but between orders and the shop floor she never wrote. With a system like this, every Monday she receives two drafts based on her customers' real questions ("can this casserole go on an induction hob?"). She goes over them with her coffee, changes four sentences and approves them. It takes her twenty minutes a week, and she has not missed a single week in eight months. The detail that makes it credible: the texts come from her answers, not from an AI inventing cooking advice.
What to watch (and it matters)
Three things I see done badly often:
- Publishing without reviewing. AI can get facts, prices or claims about your sector wrong. Everything that goes out under your name must have been read by a person from the business. Always.
- Factory tone. If the texts could belong to any company, they are no use. The system has to learn from how you speak, with your examples and your expressions.
- Where your data ends up. To write well, the system works with the business's material: customer questions, services, prices. If you use generic free tools, that information can end up training third-party models. Always look for solutions with data inside the EU and an explicit commitment that nothing of yours is used to train anything of anyone else's.
And since August 2026 there is one more thing to keep in mind: whoever holds editorial responsibility for what is published is what the EU AI Act's transparency rules turn on. I went through it in the piece on Article 50.
Start small
No grand strategy is needed to begin: a list of ten customer questions and a commitment to one piece a week already put you ahead of most competitors. The rest is a system that holds the cadence for you — with your tone, your final review, and your data always under control.
This article is general information, not advice for a specific case.
About this article
iaetica.cat is a space for writing in the open: what I am learning about artificial intelligence, its ethics and its law.
— Laura Graells