Google CTR benchmarks vs your CTR: the gap is the whole story
Every SEO post quotes the same CTR-by-position numbers. They are useless on their own. Pair them with your actual CTR per query, push the chart into Slack, and the title-tag work writes itself.
Google CTR benchmarks vs your CTR: the gap is the whole story
Every "what's the average CTR for position 3?" article cites approximately the same numbers. Position 1 gets around 28%. Position 2 around 15%. Position 3 around 10%. Then it falls off a cliff.
These numbers are fascinating in a vacuum and useless in isolation. The benchmark on its own doesn't tell you what to do. The benchmark next to your actual CTR for the queries you actually rank on — now that's a to-do list.
The chart, in three flavours
The scatter, for the technically inclined
A scatter plot. X-axis: position. Y-axis: CTR. One dot per query. Dot size: impressions. Draw a benchmark curve through the middle.
Now: anything well below the curve at high impressions is a title-tag candidate. Anything well above is a learning to apply elsewhere. The whole report fits on one chart. Scatter plots in Slack are covered in scatter charts from Sheets to Slack.
The bar chart, for everyone else
If your team finds scatters fiddly (fair), do the simpler version: a bar chart of (benchmark CTR − actual CTR), top 20 queries, above some impression floor. Highest bars are your biggest title-tag opportunities.
This is the chart you actually want in #seo. It's a queue, sorted by leverage. Bar chart specifics in bar charts from Sheets to Slack.
The trend, for everyone
A simple line of site-wide average CTR over the last 12 weeks. Doesn't surface individual queries, but tells you whether the SERP landscape is changing under you (a new SGE module appearing, a featured snippet you used to own going somewhere else). Big shifts in this line warrant a deeper look.
Why the benchmark doesn't have to be perfect
The benchmark curve doesn't need to be the right one. It needs to be a one. A rough "expected CTR by position" line pasted into the chart is enough to surface the queries that are genuinely off-trend. Use one of the public studies. Use your own historical site-wide average. It barely matters — the gap is what the eye reads.
This is one of those cases where a pretty good chart in Slack on Monday beats a perfect chart in Looker that nobody opens.
Where to send it
#seo, weekly, not daily. CTR is noisy day-to-day at the per-query level. Weekly aggregates smooth that out and match the cadence at which a human is actually willing to rewrite a title tag.
Setting it up
- Add Google Search Console as a source in Chartcastr.
- Pull the queries report with position, impressions, clicks, CTR.
- Either chart directly, or pipe it into a Sheet for the benchmark overlay and post the Sheet's chart.
#seo, weekly.
