How to Prioritise High-Impression, Low-Click Pages
Sort by impressions, filter to low click-through, then split the list in two before doing any work: pages whose queries match what the page offers, and pages whose queries do not. Only the first group has a presentation problem worth fixing with a title and description rewrite. The second group is ranking for the wrong queries, and rewriting its title to chase clicks makes the mismatch worse — those pages need either different content or the acceptance that the impressions were never yours. Check position before either: a page averaging position fourteen has a low click-through rate because it is on page two, not because of its title.
High impressions with low clicks looks like free upside, which is why every audit surfaces it. Most of the list is not upside. Three different situations produce the same shape in the report, and they need opposite responses.
The three situations
| Situation | Signal | Correct response |
|---|---|---|
| Poor presentation | Good position, queries match the page | Rewrite title and description |
| Query mismatch | Any position, queries do not match | Change content, or accept and move on |
| Just ranking low | Average position 11+ | This is a ranking problem, not a CTR problem |
Check position before anything else
Click-through rate falls steeply with position, so a page averaging position fourteen will have a low rate regardless of how good its title is. Rewriting it produces no measurable change and consumes a cycle. Filter to pages in the top ten before you call anything a presentation problem — that single filter removes most of a typical opportunity list.
Read the queries, not the page
For each surviving page, open its query list and ask whether the page genuinely answers those queries. A pricing page receiving impressions for a how-to query has a mismatch, and no title rewrite fixes it — it will either keep not earning clicks or earn clicks from people who bounce, which is worse for you and for them.
What a good rewrite changes
Specificity, mostly. A title that names what the page actually gives — a number, a format, a constraint — outperforms a generic one at the same position. Descriptions matter less than people assume, because they are frequently rewritten by the search engine, so put the effort into the title.
Measuring whether it worked
Compare the same page's click-through rate over equal periods before and after, at a similar average position. If position moved, the comparison is not clean and you cannot attribute the change to the rewrite. This is the step that gets skipped, and skipping it is how teams accumulate confident beliefs about title formats that have never been tested.
FAQ
What click-through rate should I expect? — Compare against your own pages at similar positions rather than a published benchmark. Benchmarks aggregate query types and intents that behave nothing like yours.
Are impressions without clicks harmful? — Not directly. They are a signal about what a search engine thinks your page is for, which is useful information even when it is unflattering.
How long before I can judge a rewrite? — Long enough for the page to re-establish at a comparable position, usually several weeks. Judging sooner means judging position noise.
Put this to work on your own website.
Fenn finds what your customers ask, drafts the articles and site fixes, and measures what ChatGPT, Claude, Gemini, Perplexity and Grok say about you — with every change waiting for your approval.