How does an SEO company measure campaign success?

Against the objective the campaign was set up to reach, on the pages it actually touched, compared, as far as it can be estimated, with what would have happened without it. A campaign is a bounded piece of work, such as a set of rewritten service pages, a technical fix across a template, or a content cluster, with a stated goal, a start, and an end. Measuring it means isolating its pages and searches, comparing them with a baseline and with similar pages the campaign did not touch, accounting for outside events, and then giving a plain verdict: met, missed, or exceeded.

A written objective comes first

A campaign can only be judged if its goal was written down before the work began, in terms that can be counted and with a date attached: organic inquiries from a group of service pages, visibility for a defined set of product searches, or the indexing of a section that was not indexed. “Improve SEO” cannot be judged, because nothing would count as failure. The baseline for those same measures has to be recorded before work starts.

Isolating the campaign’s pages

Sitewide totals mix the campaign with everything else that moved. Search Console’s guide to advanced filtering and comparison in the Performance report describes a Pages filter using a custom regular expression, which can hold the whole list of campaign URLs in one filter, and a “Doesn’t match regex” option that shows everything else. Comparing two date ranges adds a Difference column to the results. The same page notes that when you compare by page, the way clicks, impressions, and CTR are calculated can change, so a company should read those figures with that in mind.

A comparison group

Evidence that the campaign contributed to a change comes from a comparison with similar pages it did not touch. If the rewritten service pages gained impressions while comparable untouched pages stayed flat over the same weeks, that pattern strengthens the case that the campaign contributed. It does not prove it on its own: the two groups may differ in ways that matter, and only a comparison planned in advance to rule out other explanations comes close to proof. If both groups rose together, something that affected both, such as seasonality, a Google update, or a marketing push, is the more likely explanation. Choosing the comparison pages before the campaign starts keeps the choice honest.

Outside events and attribution

The company should list what else happened during the campaign: ranking updates on Google’s Search Status Dashboard, paid campaigns, price changes, press coverage, or site changes made by others. For conversions that involved several channels, it should say which attribution view it used, since a report that credits only the last visit can hide the part organic search played earlier in the journey.

The verdict

The report should state whether the objective was met, missed, or exceeded, with the numbers behind it and the limits of the evidence. A campaign aimed at inquiries is judged on inquiries, not on a rise in some other metric. If the timeframe was too short for the result to show, the report should say so and set a date to judge it again, rather than declaring success or failure early.