Before the first post is written, five decisions need to be made: where the blog lives (own domain under /blog/, subdomain or external service), how it's structured (categories, addresses, template), who supplies facts and approves content, how often posts appear, and how success is measured. The order matters: anyone who starts writing before responsibility and rhythm are sorted out ends up with a dormant blog after three posts.
A company blog rarely fails because of the writing. It fails because of decisions that were never made: nobody knows who approves content, the blog sits on a subdomain that no one links to, and after the third post it's unclear whether the effort is paying off. This page works through the five decisions that must be made before the first post, and ends with a checklist.
Decision 1: Where does the blog live?
| Option | Address | Advantage | Disadvantage |
|---|---|---|---|
| Own domain, subdirectory | firma.ch/blog/ | Posts strengthen the main domain, simplest linking | Website system must support a blog |
| Subdomain | blog.firma.ch | Technically independent of the main system | Google tends to treat it as a separate site; linking is often forgotten |
| External service | medium.com/firma or LinkedIn articles | No effort required | Visibility belongs to the platform, not to you |
| Hosted blog via proxy | firma.ch/blog/, content is hosted externally | Own domain without rebuilding the website | Dependence on the provider, clarify export beforehand |
For most SMEs, a subdirectory on their own domain is the right choice. If you don't have a system with blog functionality, a hosted blog via proxy gets you to the same place. The detailed trade-offs are covered in Blog on your own domain, subdomain or external.
Decision 2: How is the blog structured?
Three things need to be set before the first post, because changing them later only works with redirects:
- Addresses. Descriptive addresses that match the topic (firma.ch/blog/waermepumpe-kosten/), no date elements, no numbers.
- Categories. Three to six, based on services or customer questions, not internal departments. How many and which ones is covered in Categories and tags in the blog.
- Template. Every post follows the same structure: title as a question or answer, first paragraph with the answer, subheadings, a table or list, FAQ section, link to the relevant service page.
Decision 3: Who is responsible?
The decision that trips up most blogs. You need three roles, which can be filled by one person or three:
- Fact supplier. Knows the services, prices, projects. Usually management or a specialist.
- Writer. Employee, agency, freelancer or system. This role is the easiest to fill.
- Approval. Reads every post before publication and takes responsibility. This role can't be delegated.
If approval sits with management, it needs to be built into the weekly schedule: a fixed slot, for example Monday before the first meeting, with a maximum of ten minutes. Without a fixed slot, drafts pile up.
Decision 4: What rhythm?
One post a week is the pace an SME can sustain and that makes a difference with Google. Consistency matters more than the number: four posts in the first month followed by three months of nothing is worse than one post every two weeks for a year. Set the rhythm before the first post goes live, and plan out the first quarter. Editorial plan for the company blog shows how to do this in an hour.
Decision 5: What's measured?
Without measurement, after three months you'll be debating whether the blog is worth it, with no figures to back it up. Three things are set up before the first post:
- Google Search Console for the website, so impressions and clicks per post are visible. Instructions under Setting up and reading Search Console.
- A metric that matches the goal. Enquiries via the contact form, calls, appointment bookings. Today's figure is noted down so there's a benchmark.
- A fixed slot every four weeks to review the figures. Twenty minutes is enough.
What a blog can and can't achieve is described, with examples, in What does a company blog achieve?
Checklist before the first post
- Blog sits on the company's own domain or is connected there via proxy
- URL structure and three to six categories defined
- Template for the structure of each post is ready
- Fact supplier, writer and approver are named
- Fixed time for approval in the weekly schedule
- Rhythm set and first quarter planned
- Search Console set up, baseline for queries noted
- Three to five posts prepared for launch
- Blog is linked from the main navigation and from the service pages
Mistakes we've seen in over 90 projects
- The blog isn't linked anywhere. It exists under /blog/, but neither the navigation nor the service pages lead there. Google finds it late, visitors never.
- Approval has no deadline. Drafts sit for two weeks, then three, then the rhythm is gone.
- Topics from an inside view. "Our new delivery van", "Company outing review". Nobody searches for that.
- Twenty posts at launch, then silence. The blog looks active but isn't. Google notices the standstill.
- No measurement. After six months nobody knows whether the blog has brought in enquiries, and it gets stopped on a gut feeling, often just before it starts working.
All five mistakes can be avoided before the first post. None of them has anything to do with the quality of the writing.
Frequently asked questions
Do I need a new website for the company blog?
Usually not. WordPress, Wix and Shopify all have a blog function that can be used under the existing domain. Only if the website runs on a system without a blog do you need an addition, for example a hosted blog connected under your own domain via proxy.
How many posts should be online at launch?
Three to five, so the blog doesn't look empty and Google has several pages to work with. After that, the fixed rhythm applies. Twenty posts at once bring no advantage, they just push back the moment the rhythm begins.
What does it cost to set up a company blog?
Setting it up within an existing system is often done in a few hours. The ongoing costs depend on who writes: your own time, an agency, a freelancer or a blog system. Price ranges for these options are summarised under What does an AI blog cost.
How long should I keep the blog going before deciding whether it's worth it?
At least six months at a weekly rhythm. Experience shows the first movement in Search Console after six to eight weeks and noticeable enquiries usually from the fourth month onwards. If you stop after eight weeks, you've paid without waiting for the result.