Optimising a blog post for Google means checking it in a fixed order after writing: the question in the title, the answer in the first sentences, a subheading per sub-question, two to four internal links, an FAQ section, descriptive alt texts and a page title with a description. Counting keyword density, on the other hand, achieves nothing. Google judges content by usefulness, not by how often a term appears.
The text is written, the file sits in the CMS, and now comes the part most people skip: the check before publishing. It takes a quarter of an hour and often decides whether the post brings in visitors in three months or gets lost in the "News" category. The list below is deliberately set out in this order, from the most important point to the finishing touches.
1. Is the search question in the title?
The title must contain the question someone types into Google, in their own words. Nobody searches for "News from the workshop". "How much does a made-to-measure fitted wardrobe cost?" is found by anyone wanting to know exactly that. Check: would I type this heading as a search query myself? If not, rephrase it. If you work from a brief, you'll have settled this question before writing. The process is explained under Writing blog posts with AI.
2. Does the answer come in the first three sentences?
Scroll to the top. If there's an introduction about the importance of the topic, delete it. The first paragraph answers the question from the title so concisely that it could stand alone as a quote. Everything else, background, exceptions, examples, follows after that. Google likes to show exactly this paragraph as a snippet, and readers who find the answer at the top are more likely to keep reading than those who have to search for it.
3. Does every subheading carry a sub-question?
Subheadings aren't decoration, they're the table of contents for readers and search engines alike. Check each heading: does it say what the section answers? "Other aspects" says nothing. "What the Thurgau funding scheme covers" says everything. Use only H2 and H3 levels, in a logical order, without skipping. One section per sub-question, no longer than around 150 words, otherwise split it up.
4. Are two to four internal links in place?
Every post needs links in both directions. From the new post: at least one link to the matching service page and one to three links to related posts, each with anchor text that describes the destination. To the new post: at least one older post should link to it, otherwise the new page carries no weight. How weight flows between pages is explained on the page about internal linking.
5. Is there an FAQ section with genuine follow-up questions?
Three to five questions that someone would genuinely still have after reading: "Does this also apply to rented flats?", "What if the façade has cracks?". Each answer in two to four sentences, without repeating the main text. These questions cover search queries too small for their own post and give Google further phrasings on the same topic.
6. Do the images have alt texts and descriptive file names?
Every image gets an alt text describing what's shown: "Oak fitted wardrobe under a sloping ceiling", not "Image1" and not a string of keywords. The file name should say the same thing in short form. Also check the file size; a photo of several megabytes slows down the whole page. Details are under Image SEO.
7. Are the page title, description and address correct?
The page title (title tag) can differ from the blog title: shorter, with the search term at the front. The description answers in two sentences what the post offers. The address (URL) is short and consists of the three to five words of the question, without a date and without filler words. Formulas and examples for this are under Page title and meta description.
8. Is the date honest and the category set?
Publication date visible, category appropriate, no ten tags. If you update the post later, only change the date if something has genuinely changed in the content.
The checklist at a glance
| Step | Check question | Where it's handled |
|---|---|---|
| Title | Would I type this into Google? | Heading |
| Answer | Do the first three sentences answer the question? | First paragraph |
| Subheading | Does every heading say what the section clarifies? | H2 and H3 |
| Internal links | Two to four outgoing, at least one incoming? | Text, older posts |
| FAQ | Three to five genuine follow-up questions? | End of post |
| Images | Alt text, file name, size? | Media library |
| Meta | Page title, description, URL? | Page settings |
| Date, category | Honest and clear? | Page settings |
What you can safely leave out
Counting keyword density. This metric dates back to a time when search engines counted words. Today, Google assesses whether a page usefully answers the question, and its own guidelines explicitly state that text written for search engines achieves the opposite. Equally dispensable: bolding terms so Google 'sees' them, tag clouds, keywords in the meta keywords field (Google has ignored these for years), and aiming for a specific word count. If you're wondering whether a text will be recognised as AI-written and penalised: Google doesn't penalise AI texts across the board, but rather useless texts, regardless of who wrote them.
What to do with this
Print out the table or save it as a template in the CMS and go through each post before clicking 'Publish'. If you publish several posts a week, it's best to delegate points 3 to 7 to the system that delivers the texts: BlogPilot, for example, already creates subheadings, an FAQ section and internal links within the post, while the editorial check of the claims stays with the business. For checking the content, whether figures, names and statements are correct, there's a separate checklist before approval.
Frequently asked questions
How often should the keyword appear in the blog post?
There's no target number. The term belongs in the title, main heading, the opening sentences, and wherever it fits naturally in the text. If you count, you usually write worse, because the text gets built around the term instead of the question. Synonyms and related terms help Google more than repetition.
Should I optimise the article before or after writing it?
Both, but different things. Before writing, you define the search query, intent and outline. After writing, you check structure, links, FAQ and meta details using this list. If you optimise while writing, you lose the thread and the reader.
Does optimisation also apply to AI texts?
Yes, the list applies regardless of who wrote the text. According to its own statements, Google assesses the usefulness of a text, not its origin. AI texts arguably need the check even more urgently, because they often put a long introduction before the answer and forget internal links.
How long does the check take per post?
With a bit of practice, 15 to 20 minutes. The internal links take the longest, because you need to find suitable older posts and add a link back there. If you keep a list of all posts with topic and address, you'll be done in ten minutes.