All articles
Case Study
5 min readOctober 30, 2026

Case study: an e-commerce site was overestimating CTR by 3x for two years

An e-commerce brand reported a 12% CTR in stakeholder decks for two years. The real number was 4.1%. How the discrepancy went unnoticed and what happened when accurate data replaced the inflated metric.

For two years, the SEO team at a mid-market e-commerce brand presented a monthly deck showing a CTR of 12%. The number came directly from Google Search Console: total clicks divided by total impressions from the keyword export. It matched what GSC displayed. Nobody questioned it.

How a 3x overstatement went unnoticed

The keyword export showed roughly 30,000 impressions and 3,600 clicks per month. The CTR was 12%. But SearchVault's sentinel-row analysis revealed the property's true impression total was 88,000 per month. The real CTR was 4.1%.

Why did nobody catch it? Because the number was consistent. Each month the keyword table showed approximately the same proportion of hidden impressions, so the inflated CTR trended smoothly. Trends were real; the baseline was wrong. The team was watching a correct trend on an incorrect axis.

The impact on decisions

  • Content ROI models assumed a 12% CTR, making every new page look 3x more profitable than it actually was.
  • Budget allocation favoured SEO over paid search partly because organic appeared to convert attention at an unusually high rate.
  • Competitor benchmarking showed the brand outperforming industry averages that were computed from complete data, a comparison that made no sense.

The correction

The team rebuilt their reporting on sentinel-aware CTR. The board initially questioned the "drop" from 12% to 4.1%, until the methodology was explained: the old number excluded 66% of impressions. Once the baseline was corrected, two things happened.

  • The SEO team identified 58,000 monthly impressions coming from queries they had never optimized for, because those queries had been invisible in the keyword table.
  • Three product categories showed CTR well below the property average in the complete data. Those became immediate optimization targets and delivered a genuine (not arithmetically inflated) CTR improvement of 1.3 percentage points over the next quarter.

The full methodology is described on the methodology page. If you suspect your own CTR may be overstated, try the free CTR calculator to get an instant estimate of your data accuracy gap.

The most dangerous number in SEO reporting is the one that looks consistent. A stable CTR computed from a truncated denominator is consistently wrong.

See it in action

Suspect your own CTR is overstated? Try the free CTR calculator to get an instant gap estimate from your own CSV, or sign up for SearchVault to see the sentinel-aware numbers for every property in your portfolio.

See it with your own Search Console data

Request access and we will set up an account built around the properties you already own.