Skip to content
Serplyze
Start free
All docs

Content findings

How Serplyze finds losing keywords, page-two keywords, CTR gaps, cannibalised keywords and device gaps, and how clicks at stake are estimated.

Content turns your tracked keywords into a work list. Five checks run over their Search Console data, and each finding carries an estimate of the clicks a month at stake, so you can start with the largest.

By default the checks use the last 28 final days, and declines are measured against the 28 days before. The calendar sets a custom period and a comparison (Previous period, Same period last year or Custom). Provisional days are never used.

The five checks

CheckA tracked keyword is listed when
Losing keywordsThe comparison period had at least 10 clicks or 100 impressions, the keyword lost clicks, and either clicks fell by 20% or more and by at least 10, or the position got worse by 1 or more with fewer clicks and 100+ impressions before.
Page twoIts average position is above 10 and at most 20, and it has impressions.
CTR gapIts position is 10 or better with at least 50 impressions, and its CTR is under 70% of the expected CTR and at least 1 percentage point below it.
CannibalisedIt has at least 20 impressions and two or more of your URLs each earn at least 10% of its impressions (or of its clicks, once it has 5 or more clicks).
Device gapMobile and desktop each have at least 100 impressions, the stronger device ranks in the top 20, and either the positions are 3 or more apart, or at a similar position one device is clicked at less than half the other's level.

Losing keywords need at least 7 days in each window. A finding is listed only when its clicks at stake round to at least one click a month.

Plan

Cannibalised and the Pages view use the ranking pages of each keyword, which are stored on paid plans only. The other four checks run on every plan.

How clicks at stake are estimated

Each estimate is computed over the period and scaled to a 30-day month (Cannibalised is the exception: it shows the period's clicks as they are).

CheckClicks at stake
Losing keywordsClicks in the comparison period minus clicks now
Page twoImpressions × expected CTR at rank 8, minus current clicks
CTR gapHalf of (expected CTR − CTR) × impressions, capped at twice the current clicks (counted as at least 10)
Cannibalised“Clicks to others”: the clicks of every competing URL except the leading one
Device gapHalf of what the weaker device would gain at the stronger device's CTR level, with the same cap as the CTR gap

The summary at the top, Clicks within reach / month, counts each keyword once, at its largest estimate, and lists the five findings with the most at stake.

Expected CTR

CTR gap and Page two do not compare your keywords with an industry average alone. The expected CTR starts from a generic CTR-by-position curve and is scaled to how your own keywords are clicked:

  • Only non-brand tracked keywords in the top 20 with at least 20 impressions count toward the baseline.
  • They are pooled by position bucket (1, 2, 3, 4–5, 6–7, 8–10, 11–20), weighted by impressions. No keyword weighs more than 25% of a pool, and the keyword being judged is left out of its own baseline.
  • With few keywords the baseline is pulled toward the generic curve.
  • When the keyword is in a keyword group with at least 3 other eligible keywords, the group's own CTR level is used. A cluster of queries that Google answers on the results page then sets a lower bar.
  • The scaling factor is kept between 0.05 and 1.5.

How this is estimated in the side panel names the baseline used for the selected keyword.

Working a finding

  1. Click one of the five cards to list its findings, largest first. Filter by column, or group the table by keyword group.
  2. Click a row. The side panel shows the facts, a chart of position and CTR over the last 90 days, a recommendation, the pages ranking for the keyword and the CTR of similar keywords in the same group.
  3. Make the change, then click Log action in the panel. The action is saved with the keyword filled in and the finding is marked done.
  4. Or use Mark as done and Ignore directly. The tabs Open, Done, Ignored and All filter by status.
  • A finding marked done that is still found 28 days later shows as open again.
  • Ignored findings stay ignored until you click Un-ignore, and are left out of reports.
  • On the Done tab, the Result column compares the keyword's clicks 14 and 28 days after the fix with the 28 days before it, against the whole site's trend. A change within ±10%, or with fewer than 20 clicks, reads “No clear change”.
  • Cannibalised rows have no status; they open the keyword page with all of its ranking URLs.
  • The View switch shows open findings by Keywords or by Pages (the page each keyword lands on).
Limit

The checks cover tracked keywords only. For every query on the site, use the Query × page tab and the quick views in Traffic. Clicks at stake and fix results are estimates, not measurements.