How to Start a Blog That People Actually Read

Starting a blog takes an afternoon. Getting anyone to read it takes longer, and most of the advice about it is written by people selling something. Here is the order of operations that actually works, with nothing to buy.

Pick a subject narrow enough to be useful

"Technology" is not a subject. "Migrating small Rails apps off Heroku" is. A narrow subject gives you two things at once: readers who were searching for exactly that, and a clear sense of what to write next.

The test is simple. Can you list ten article titles right now? If you can, the subject is narrow enough. If you are staring at the ceiling, narrow it further.

Write the five articles you wish had existed

Think back to the last time you had to work something out the hard way. What would have saved you three hours? Write that. Those articles are useful by construction, because you already know they were missing.

Do not start with a "welcome to my blog" post. Nobody reads those, including you.

Make each article finishable

• Answer the question in the first paragraph. Readers who need the short version get it and leave happy; that is a good outcome.
• Use headings that describe the content, not clever ones. People scan before they commit.
• Keep paragraphs to three or four lines. A wall of text reads as work.
• Show the actual thing — the command, the screenshot, the number. Vague articles are forgettable.

Publish on a schedule you can keep when you are busy

Once a fortnight for two years beats three times a week for a month. Pick the pace you could keep in your worst week, not your best one.

Then, and only then, worry about traffic

Search engines need something to rank. Ten solid articles give you ten chances; one gives you almost none. Once the articles exist:

1. Write a specific title and meta description for each one.
2. Submit your sitemap to Google Search Console and watch which queries bring people in.
3. Write more about whatever those queries are, because that is your audience telling you what they want.
4. Link your own articles to each other where it genuinely helps the reader.

The best time to start was a year ago. The second best time is this afternoon, with an article you wish someone had written for you.

What to ignore

Ignore keyword density, ignore posting frequency targets you cannot sustain, and ignore anyone promising guaranteed traffic. Write things worth reading, publish them consistently, and give it a year.