SEO PlaysFind low-competition keywords by reading the SERP, not the score
A keyword difficulty score is mostly a link count. It looks at the pages ranking now, weighs the domains pointing at them, and returns a number — a fair estimate of how hard those pages are to out-link, and no estimate at all of how hard they are to out-write.
Those are different questions, and the score can only answer the first. It cannot see that four of the top ten are forum threads, that the leading result has not been touched in years, that nobody has written a page matching what the searcher actually wants. Two minutes of looking at the results yourself sees all of it.
So you look, and you write the verdict down. Ten minutes a keyword, four things noted per result, and a sheet you check against reality two months later — which is the part that turns a habit into judgment.
What you get
A repeatable ten-minute hand-check of whether you can win a keyword this year, catching the winnable SERPs difficulty scores miss.
Steps
- Build a candidate list of 30 to 50 phrases from autocomplete, People Also Ask, Reddit threads, and competitor pages.
- Google each candidate in a clean browser window (private mode, or a profile with no history), so personalization is not deciding what you see.
- For each top-ten result, note four things: how big the brand is, whether the title matches the query exactly, how old the content looks, and what format it is.
- Green flags: Reddit, Quora or forum threads in the top five; pages two or more years old; results that only partially match the intent; no exact-match titles.
- Red flags: all ten results are recent, dedicated pages from brands with real authority. Close the tab; the score was right this time.
- Count the results you could beat on content alone. Write the page only where that count is three or more.
- Match the format the SERP already rewards — list, guide, tool, comparison — instead of the format you prefer; the formats in the top ten are the ones this query has rewarded already.
- Keep a sheet of every verdict, and check it against reality in four to eight weeks to calibrate your judgment.
How to verify it worked
- Search Console shows impressions for the chosen keyword within two to six weeks of publishing.
- Your verdict sheet shows a rising hit rate as the calibration loop runs.
Why it works
The four things you note per result are the four ways a ranking page is beatable without links: a small brand, a title that does not match, aging content, the wrong format. Counting how many of the ten fail on at least one of those is the entire judgment, and three is the line worth writing for.
Evidence
The second-most-repeated tactic in a year of practitioner posts we distilled (49 separate posts): check authority, backlinks, age and intent by hand before writing anything.
Ahrefs, one million new URLs: only 1.74 percent reach the top ten within a year — and of those that do, 40.8 percent get there within a month. Target picking is most of the outcome.
Our own SERP pass for this site found KD-0 queries whose top ten was low-authority listicles and forum threads — the shape this check exists to catch.
Last reviewed: 2026-09-01. A play whose evidence stops holding up is rewritten or retired.