Striking distance keywords: which ones are actually worth chasing
Everyone tells you to find keywords ranking 4 to 20 and push them. Nobody tells you which ones repay the effort.
Short answer
Export your queries, keep non-brand rows between positions 4 and 20, and sort by impressions. Then size each one: impressions multiplied by the CTR of the band you are aiming at. On this dataset, moving a page from the 8 to 10 band into 3 to 4 multiplies its clicks by roughly fifteen without a single extra impression.
The filter
Non-brand only, position between 4 and 20, sorted by impressions descending. Brand queries have to go first or they dominate the list with terms you already win. Position 4 to 20 is the band where the page is understood and simply outranked; below 20 the problem is usually authority, not the page.
Sizing the payoff properly
Take the impressions a query already earns and multiply them by the CTR of the band you are aiming at, measured rather than borrowed: 0.81% at positions 8 to 10, 9.09% at 4 to 5, 11.82% at 3 to 4. Moving one query from the bottom band to the top one multiplies its clicks by roughly fifteen, on the same impressions. That ratio, not the volume, is the reason to pick one query over another.
The ones to skip
Queries with high impressions and a position below 30: those are page-three appearances and no amount of on-page work moves them. Queries with tiny impressions at a good position: already won, nothing left. And anything where the intent does not match what your page does, however close the ranking.
What actually moves one
At position 4 to 10 the usual cause is that a better page exists, not that yours is broken. Completeness, specificity and speed are the three things you control. If those are already right, the honest answer is authority, and that is a different project with a longer timeline.
Questions
- How long before a striking distance fix shows up?
- Roughly a quarter. On this dataset the month that shipped nothing grew 61%, because the work that earned it had shipped twelve weeks earlier.
- How many should I work on at once?
- Few. The top five pages on this site earned 75.2% of all clicks, so concentrated effort on a handful beats spreading it across fifty.
Measured, not asserted
Free tool for this: Search Console Analyzer. No account, nothing uploaded.
Where this goes deeper
Every number on this page comes from one complete dataset: one product taken from zero to 250K+ users on search alone, with nothing spent on advertising. The full argument is chapter-20-averages-describe-nobody of the book. Five chapters are free to read.