The short answer

A blog is worth it because every good article answers one more of your customers' questions. That opens one more door to your website, on Google and in AI search like ChatGPT. From our project experience, one thorough article a month is enough to start. What matters is that it answers a real customer question better than the pages that show up for it today.

Why aren't a few static pages enough?

Many SME websites consist of five pages: home, services, about us, contact, legal notice. Each of these pages covers one topic. The service page can be found for "web design Stuttgart". But it isn't built for "What does a website relaunch cost?" or "How long does it take?".

These are exactly the questions customers ask early, long before they enquire. Whoever helps there is remembered when the decision is made. A blog extends your website with these questions. Each article takes on one topic, instead of the homepage having to rank for ten topics at once.

Here's what that looks like in practice. Example (fictional): a painting business from Ludwigsburg.

Customer's questionMatching articleLinks to
What does it cost to have a flat painted?Cost article with ranges from own jobsService page interior painting
Which paint is suitable for a bathroom?Guide with an honest recommendation per situationService page bathrooms and wet rooms
When in the year is it best to paint the façade?Guide with rules from practiceService page façade
Do I have to clear the room before painting?Short guide to preparationContact with appointment request

None of these questions fits on a service page. But each of them is a moment when a potential customer is looking for a business right now.

Why does a blog also count for ChatGPT and Google AI Overviews?

AI search systems answer concrete questions: "What does X cost?", "How long does Y take?", "What should I watch out for with Z?". A website without content on these questions has nothing that can be cited. That also applies to a beautifully designed homepage.

Google describes in its own guide how its AI features work [1]. They fetch suitable pages from the search index via the normal ranking systems. The model can ask several related sub-questions at the same time ("query fan-out"). A page that answers one of these sub-questions well can make it into the answer that way. That's exactly what a blog is built for.

Google also stresses that interchangeable content brings little [1]. That means texts that only repeat what's written everywhere. What's wanted is your own experience and assessment. More on the differences between classic search and AI search is in the article SEO vs. GEO.

How do you recognise a good article?

  • One question per article. Several loosely related topics in one text water down the answer.
  • The answer comes first. The first paragraph answers the question. Details, exceptions and examples follow.
  • Concrete figures and examples. "Usually between €4,000 and €15,000" helps more than "it depends". Label your own values as your own experience; other people's figures get a source.
  • A real author. Google explicitly recommends author information where readers expect it [2]. A name with a photo and background builds trust.
  • Honest limitations. A paragraph "When this doesn't fit" shows you don't only want to sell.
  • Real updates. Google even names a new date without new content as a warning sign [2]. Update facts, not the timestamp.

How is a good article structured?

You don't have to reinvent the wheel every time. This outline works for most guide articles. The example is again the fictional painting business with the question "What does it cost to have a flat painted?":

  1. Headline = search question. The way customers would type it.
  2. The short answer. Two to four sentences with the key statement, for example a price range from your own jobs.
  3. What influences the price. Area, condition of the walls, paint, furniture. Each point with a sentence of explanation.
  4. A worked example. A typical three-room flat, step by step.
  5. What you can do yourself. Clearing out, masking, planning dates. That shows you don't only want to sell.
  6. When it gets more expensive. The honest limitation, for example with mould or old wallpaper.
  7. The next step. A link to the service page and a clear button for the enquiry.

It's best to phrase subheadings as questions. That way readers quickly find their part. A table of contents at the start also helps with longer texts.

How do you find topics people really search for?

The best pool of topics is rarely your own imagination. It lies in what customers ask you anyway. You can tap these sources right away:

  • The last ten first calls. Which questions came up again and again? Which worry stood before the decision?
  • Quotes that didn't go through. Often it fails because of a question that was never answered well.
  • Emails and calls. What do customers ask after buying? Those become guides.
  • Google itself. The suggestions while typing and the "People also ask" box show how people phrase things.
  • Google Search Console. The Performance report shows queries for which you're already shown but hardly get clicks. Those are often the quickest wins.

A simple routine: collect 20 questions in a table. Mark the ones close to a buying decision. Start with these; basic topics come later.

How often should you publish?

Consistency beats frequency. One article a month that answers a question thoroughly brings more than four superficial ones. That's our project experience, not a fixed rule.

More important than the cadence is the purpose of each article. It should either cover a new question or noticeably improve an existing article. A second article on the same topic competes with the first. Better to expand the old text.

Be careful with volume. Google treats many pages that are created mainly for rankings rather than for users as spam. Pages generated with generative AI without added value are named explicitly too [3]. Ten AI articles a week are therefore not a plan, but a risk.

How do you connect blog, service pages and LinkedIn?

An article that stands alone is worth less than one in a network. Link every article to the matching service page. Conversely, link from the service page to the most important articles. That way readers find the next step, and search engines understand the connection.

Our knowledge section is built like this. One overview page per topic, several in-depth articles, and links in both directions between knowledge and offer.

Then there's LinkedIn. A good article is raw material for several posts: a figure, a list, a mistake from practice. LinkedIn brings reach in your network, and it does so now. The blog stays findable as long as the content is current. Together they make a content strategy, not an either-or. How to become visible on LinkedIn with your own face is in the article Social selling and personal brand.

How do you measure whether your blog works?

A blog works slowly. In our experience you see the first signals after a few weeks, robust trends rather after months. Three metrics are enough to start:

  • Impressions per article. The Search Console Performance report shows which queries an article appears for. If impressions rise, Google is interested.
  • Clicks to the service page. In your web analytics you see how many readers go on from the article to the offer.
  • Enquiries with their origin. Ask in the first call how someone came across you. This one question is often more honest than any statistic.

Pure visitor numbers matter less. An article with 50 matching readers a month can bring more than one with 5,000 random ones. Which reports also show your visibility in AI answers is in our overview SEO & GEO.

When isn't a blog worth it?

A blog isn't a must. In three cases we'd rather advise against it:

  • Nobody has time for their own content. Without your own experience you only produce general knowledge. Nobody reads it, and it doesn't get cited either.
  • Your customers don't search online. If your business runs almost entirely on referrals and tenders, other channels matter more.
  • The website has bigger problems. If the homepage doesn't bring enquiries, start there. A blog amplifies a working website; it doesn't repair a broken one.

Whether your website is ready is shown by the 7 signs your site needs an update. For planning a new site, the website project checklist helps.