Help & Documentation
Browse the full documentation index · Press Esc to close
Breakdowns

CTR Curve

CTR Curve plots the click-through rate your site actually achieves at each rank position on the search results page, computed from your own 90 days of Search Console data. The shape of the curve — how steeply CTR falls from position 1 to position 10 — is one of the most useful diagnostic tools in SEO. Once you know your own curve you can stop relying on industry-average benchmarks and instead say with confidence "a query at position 6 with 1.5% CTR is performing in line with our site" or "below it".

What it measures

  • Position bucket — integer rank position from 1 down to 10 (and beyond, bucketed in the long tail).
  • Impressions / Clicks — totals over the trailing 90 days at that rank.
  • CTR — clicks ÷ impressions for the bucket, the headline number on the curve.

How we compute it

  1. Across the trailing 90 days, every query × URL impression carries an average position. We round that position down to an integer bucket (so position 3.4 contributes to bucket 3).
  2. For each bucket we sum impressions and clicks, then compute CTR as the ratio.
  3. The result is a curve: bucket 1 typically sits in the 25–35% CTR range; bucket 10 typically sits below 1%; and the slope between them is your site's reality.

Why a custom curve, not an industry benchmark?

Industry-average CTR curves (the ones you see in blog posts) are computed across thousands of sites and millions of queries. They're useful as a sanity check but wrong for any individual site, because the real CTR at any rank depends on: how branded vs unbranded your queries are, how much SERP-feature density (PAA, AI overviews, ads, image packs) is interleaved above your link, and how compelling your title and meta description are. A site-specific curve captures all of those effects automatically.

Scenarios you'll see

Above-curve performer

A specific query in CTR Opportunities outperforms the bucket average — strong title, brand, or rich result. Worth studying as a template you can copy onto under-performers in the same band.

Below-curve underperformer

A query whose CTR is materially below your own bucket average is the textbook candidate for a title/meta-description rewrite. The CTR Opportunities report uses your curve as the implicit benchmark.

AI-overview dampening

If your bucket-1 CTR is well below the typical 25–35%, AI overviews or featured snippets are intercepting clicks. Your curve absorbs this — that's why you should benchmark against your own curve, not an industry one.

Branded skew

Sites with strong brand recognition often show a very steep curve (very high bucket-1 CTR). That's expected — and a reason not to be alarmed when non-branded queries on your site achieve lower bucket-1 CTRs than the curve suggests.

Unusual flat zone

If buckets 4–7 all have similar CTR, the SERP for that band is noisy with features. Less ranking-improvement leverage; consider focusing on getting into bucket 1–3.

Long-tail cliff

CTR collapses below 0.5% past position 10 — confirms why Striking Distance focuses on positions 11–20: that's the band where a small ranking gain produces an outsized CTR gain.

What to do with it

  1. Use this curve as your benchmark when reading CTR Opportunities and Top Queries — query CTR is "good" or "bad" relative to your own curve, not an industry average.
  2. If your bucket-1 CTR is unusually low (< 15%), audit the SERPs of your top queries: AI overviews / PAA / featured snippets are likely eating clicks. Lean into schema and brand.
  3. Re-check the curve after major SERP shifts (Google updates, AI overview rollouts) — the curve will adapt and so should your expectations.

Caveats & limits

  • Position is reported as an average per query × URL × day. The integer bucketing is therefore a smoothing approximation, not a literal "ranked first today" count.
  • Sites with very low traffic produce noisy curves; the curve becomes meaningful with at least a few thousand impressions across the buckets.
  • The 90-day window means very recent SERP-feature changes take a few weeks to fully reshape the curve.

Related reports