Keyword research for a maker blog: how I picked what to write about

When I started writing notes for this site, I did what most engineers do: wrote about whatever I'd just finished. Some of those posts are good. Almost nobody would ever find them. So for later notes I added a light layer of keyword research — enough to aim, not enough to ruin the writing.

Start from what you actually know

Keyword research works best in reverse. List the things you've genuinely done — for me: ESP32 boards, KiCad layout, ordering PCBs from JLCPCB, 3D-printed enclosures, LoRa radios, sales funnels, email automation, backtesting a stock scanner. Then ask what someone searching for each of those would type.

"Board bring-up checklist" is a search. "My thoughts on Rev C" is not.

Use a tool to check the guesses

Tools like SE Ranking and Semrush answer three questions quickly:

1. Volume — does anyone search for this phrasing?
2. Difficulty — who already ranks, and could a small site compete?
3. Related questions — what else do people ask around the topic?

For a niche technical blog, low-volume, specific phrases are the sweet spot. "ESP32 deep sleep current" will never be a big keyword, but the people who search it are exactly the readers you want.

Then track whether it worked

Rank tracking closes the loop. A note that climbs to page one for its phrase stays as is. One that stalls gets a better title, a clearer opening paragraph or a section answering the question people actually asked. Without tracking, you're guessing twice.

Rules I keep

• One main question per note. If a note answers three questions, it ranks for none.
• Title says what the reader gets. Clever titles are for people who already know you.
• Write the answer in the first screen. Then explain it.
• Never write a note only for a keyword. If I haven't done it, I don't write about it.

The goal isn't traffic for its own sake. It's making sure the person searching "why won't my ESP32 boot" at midnight finds the note that would have saved me an evening.