Skip to content

Actions and impact

Log the SEO changes you make and see a before/after comparison of the pages and keywords they targeted, set against the whole site's trend.

Actions is a log of the changes you make to a site: a rewritten title, a content update, new internal links, a technical fix. For each one, Serplyze compares the pages and keywords it targeted before and after the change, and sets the result against how the whole site moved in the same days.

Plan

Actions are included in every paid plan and are not available on the Free plan.

Log an action

  1. Open Actions in the project and click + Log action.
  2. Fill in What changed? (up to 200 characters) and Live since, the day the change went live. A planned change can be dated up to a year ahead.
  3. Pick a Type: Title / meta, Content update, Internal links, Technical, New page or Other.
  4. Add the targets: Pages (one URL or path per line, up to 20, on the project's site), Keywords (up to 200 tracked keywords) and, if you want, a Page group.
  5. Optionally set an Owner, add Details (up to 2,000 characters) and tick Visible to client.
  6. Click Log action.

Log action links elsewhere open the same form filled in: on a page detail, on a Content finding (saving also marks the finding done), on a Content Decay row, and on rows of the Traffic quick views CTR drop at top and Striking distance.

Measurement windows and lag

  • Before: the 28 days ending the day before the change.
  • After: 28 days starting 7 days after the change. The 7 lag days give Google time to recrawl and re-rank, and belong to neither window.
  • Under Measurement windows in the form you can set the Lag (days) from 0 to 60 and the Window (days) from 7 to 90. Multiples of 7 keep the weekdays balanced.
  • Page numbers come from Search Console for all queries landing on the page. Keyword numbers come from your stored keyword history.

The Measuring state

An action shows Measuring with a ready date until the last day of the after window is final in Search Console. Until then the detail panel shows an interim comparison of daily averages over the final days so far. Interim numbers are never classified.

How the result is classified

The result is judged on one target: the pages when the action has any, otherwise the page group, otherwise the keywords. An action with no target is judged on the whole site, without a baseline.

uplift = (1 + change of the target) / (1 + change of the whole site) − 1
  • The change is judged on clicks when the two windows hold at least 20 clicks together, otherwise on impressions when they hold at least 200.
  • Positive: an uplift of +10% or more. Negative: −10% or less. Neutral: in between.
  • With too few clicks and impressions, the average position decides: 1 position better is Positive, 1 worse is Negative. Position is not compared with the site.
  • No impressions before and some after counts as Positive, shown as new visibility.
  • Not enough data: no impressions in either window, or the before window is not fully covered by the stored history.

Example: the target's clicks rose 20% while the site's rose 15%. The uplift is 1.20 / 1.15 − 1 = +4%, so the result is Neutral.

The list and the detail panel

  • Filter the list by status (Measuring, Positive, Neutral, Negative, Not enough data), by type, or with the search box.
  • Click an action to open Before / after: clicks, impressions, CTR and average position for each target and for the whole site as the baseline, with small daily charts.
  • Measure again recomputes the result. Export downloads the list, or one action's before/after table and daily series.

Sharing with clients, charts and reports

  • Visible to client is off by default. When it is on, the action appears read-only in the client portal and in reports.
  • The Actions block of a report lists the client-visible actions dated or measured in the report period, each with its result.
  • Every action is drawn as a marker on the project's charts on its live date. The marker's tooltip links to the action.
Limit

A before/after comparison shows correlation, not proof of cause. The site baseline removes swings that moved the whole site, but seasonality of these particular pages, other changes made in the same weeks and Google updates can still play a part.