The content plans that get abandoned are the ambitious ones. Two thousand words a week, a video series, a newsletter, and a posting schedule across four platforms — drawn up in January with genuine enthusiasm and dead by mid-February. The plan was not wrong about what would work. It was wrong about how much time you have.
Here is a system built around a single half-hour block a week. It will not make you a publisher. It will keep your site growing, which is the part that compounds.
Where topics come from (you already have them)
You do not need topic ideas. You need to notice the ones you generate every week and throw away.
- Questions from quotes and phone calls. Every question a customer asks before buying is a page. If three people asked it, thousands are searching it.
- Objections. 'Why is it more than the guy on Marketplace?' is a genuinely excellent article, and it pre-sells for you.
- Search Console queries. Go to the Performance report, sort by impressions, and look for queries where you appear but rank badly. Those are topics with proven demand and a proven gap.
- Your own mistakes. 'Three things we got wrong on our first install' outperforms almost any polished corporate post, because it is the only kind of content nobody can fake.
- Seasonal reality. Whatever the phone rings about in October is what people search in October. Write it in September.
The four-week rotation
Different weeks get different jobs. Rotating tasks means you always know what this week's thirty minutes is for.
| Week | The job | What you finish with |
|---|---|---|
| 1 | Write one answer article | A 500–700 word answer to a real customer question, published |
| 2 | Improve an existing page | One older page updated: new FAQs, better title, refreshed price, current photos |
| 3 | Add proof | One review requested and published, or two photos of finished work added to a service page |
| 4 | Check and tidy | Search Console review, one broken thing fixed, next month's four topics chosen |
That is twelve new articles a year, twelve improved pages, twelve pieces of fresh proof, and a site that never quietly rots. It is far more than most local competitors manage.
The thirty-minute writing session
- 01Minutes 0–3: pick and frameTake the top item from your capture list. Write the question as your working headline, exactly as a customer would say it. Do not make it clever.
- 02Minutes 3–8: answer it in one paragraphWrite the direct answer first, in three or four sentences. If someone read only this, they would be helped. This paragraph is also what wins featured snippets, so it earns its place twice.
- 03Minutes 8–20: three sections of detailAdd three H2 sections that expand the answer: why it is that way, what changes it, and what to do about it. Two short paragraphs each. Resist adding a fourth section — the article is not better for it and you will not finish.
- 04Minutes 20–25: specifics and proofAdd at least one real number, one local reference, and one image of your own work. This is what makes the piece yours rather than something anyone could have written.
- 05Minutes 25–30: title, description, link, publishWrite the title tag and meta description. Link to one related page on your site, and from one existing page back to this one. Publish. Do not reread it a fourth time.
Where the leverage is: updating, not writing
The highest-return content task for most small business sites is not writing something new. It is improving a page that already ranks on page two.
- 01Find the near-missesIn Search Console, filter the Performance report to positions 8 to 20. These are queries where you are one improvement away from real traffic.
- 02Read the page as the searcherTake the query and read the ranking page as if it is the only thing you know about the business. What question does it fail to answer?
- 03Add the missing partUsually a price, a timeline, an FAQ block, or a clearer heading that matches the query wording. Twenty minutes of edits, not a rewrite.
- 04Note the date and check in six weeksMovement on an updated page is generally visible within four to eight weeks. Write the date in your notes so you can tell whether it worked rather than guessing.
What to skip
A half-hour a week means saying no to most things. These are the ones to say no to first:
- Posting the same thing on five platforms. Pick the one where your customers actually are and ignore the rest.
- Industry news roundups. Nobody local searches for these and they date within a month.
- Anything that requires a video editor. A phone-shot 45-second clip is fine; a produced series is a second job.
- Publishing AI output unedited. It reads as generic because it is generic, and generic is the one thing a local business cannot afford to be. Use it to draft an outline, then put your own numbers, photos and opinions in.
Common questions
- How often should a small business publish new content?
- Once or twice a month, consistently, beats eight posts in one week followed by silence. Search engines and customers both respond to a site that is evidently maintained; neither rewards a burst that stops.
- How long should a blog article be?
- Long enough to answer the question completely — often 500 to 900 words for a direct customer question, and 1,200 or more for a full guide. Word count targets produce padding, and padding is easy to spot.
- Is it worth blogging if nobody reads the blog?
- Judge it by search traffic to individual articles and by whether it shortens your sales conversations, not by blog homepage visits. Nobody browses a local business blog; people arrive on one article from a search, get their answer, and remember who answered it.
- Can I just use AI to write my articles?
- You can draft with it, but publishing unedited output is a competitive mistake rather than a policy one. It cannot know your prices, your photos, your local jobs or your opinions — and those are the only parts a competitor cannot copy.