For about two months, my AI team published a blog post almost every night for a wellness clinic I work with. The posts were checked, they followed the client's rules, and they went out on time. When I finally looked at how people found the site, the result was humbling. Search was sending almost nobody to those posts. Nearly every visit from search landed on the homepage.
Nothing was wrong with the writing. We had picked the wrong blog topics.
Why didn't the blog posts get any search traffic?
We chose topics from an idea bank, which is a running list of things worth saying about the business. The list was full of good material: what a first visit is like, how to prepare, why the clinic asks health questions before booking.
Those are all things the business wants to say. They are not always things a stranger types into Google.
That is the trap. A topic can be useful and true and still have nobody searching for it. You can write a great answer to a question no one asks, and it will just sit there.
- The idea bank showed us what the clinic knows.
- Search data shows what people want to know.
- Where the two overlap is where a post gets found.
How I pick blog topics now
The fix wasn't to write more, or to write better. We changed where topics come from. For this client, the rule now lives in its setup, so every future content week follows it automatically.
Step 1: Start from real searches
Blog topics come from search queries first. These are the phrases people already type that lead toward this kind of business: the service plus their town, "who should avoid it," "what to do after," "how much does it cost."
Step 2: Use the idea bank only as a backup
A topic from the ideas list with no search behind it is allowed only when the search-based ideas run out. When that happens, the topic gets flagged in the plan, so nobody mistakes it for a topic people are looking for.
Step 3: Refresh the list from what people actually typed
When the search list runs low, we refill it from the site's own search data. That data shows the exact words that brought people in. The list keeps growing out of real behavior instead of guesses.
The same idea applies to whole pages
The same week, we built pages around real searches instead of blog posts. People look for a service plus the name of their own town. So the clinic now has one page for each nearby town it serves, with honest drive times, parking details and a booking button. We also rebuilt its main "what is it" article into a full guide that answers the questions people ask most.
None of this is clever. It's simply writing for the person searching, not for ourselves.
What should a small business take from this?
If you post regularly and the visitors aren't coming, don't assume you need more posts. Check whether anyone is searching for the topics you picked.
- Look at what people already type before you decide what to write.
- Keep your ideas list, but treat it as a reserve, not the plan.
- Mark the topics nobody searches for, so you know which posts are for loyal readers and which are meant to be found.
Publishing every day is a habit. Getting found is a choice you make before the writing starts.
A blog that answers the questions people are already asking will get found long after a blog that answers the questions you wish they'd ask.