Discussion
  • Agency
  • Search Console

Year-on-year Search Console comparisons now cross the September 2025 num=100 change. Are you restating baselines, or annotating and moving on?

Published 1 reply

Discussion starter: an agency question, posted with an answer from Niraj Raut to open the thread. If you have dealt with this on a site, reply with what you saw, especially where it differs.

Disclosure: Niraj Raut, who posted this answer, runs the SEO consultancy linked at the end of it.

1 reply

  • admin
    Admin
    Original poster

    Annotate and move on for clicks, restate or avoid year-on-year comparisons for impressions and average position, and note the calendar: from October 2026, year-on-year comparisons fall entirely after the September 2025 change, so the problem largely disappears for new reporting periods. When Google stopped supporting the &num=100 results parameter, many sites saw desktop impressions fall and average position improve, because fewer automated deep result fetches were being counted. Clicks were reported to be affected far less. That makes clicks the safe year-on-year metric across the change, and impressions and position the ones that need care.

    What changed in the data

    Metric Effect of the change Year-on-year treatment
    Clicks Reported to be affected far less than impressions Compare directly, with an annotation
    Impressions Many sites saw drops, concentrated on desktop Restate, segment by device, or avoid comparisons that span the change
    Average position Often improved without any real ranking change Avoid year-on-year across the change
    CTR Often rose because impressions fell Avoid year-on-year across the change

    Options for reporting

    1. Lead with clicks. Make clicks the primary year-on-year metric for any period that spans September 2025, with a note explaining why.
    2. Segment by device. The change mainly affected desktop impressions. Mobile impressions are more comparable, so a mobile-only year-on-year comparison is a reasonable restated view.
    3. Compare post-change periods. For impressions and position, compare with the months after the change rather than with the same month a year earlier, until a full year of post-change data exists.
    4. Restate the baseline where necessary. If a client needs a year-on-year impressions figure, adjust the pre-change baseline using the drop you observed around the change, and label it as an estimate.
    5. Annotate every chart that crosses the date, including in Search Console, where custom annotations make the note visible to anyone who opens the property.

    The timing point

    As of late September 2026, the same months last year were already affected by the change. From October 2026, comparisons between months in 2026 and 2025 compare post-change data with post-change data, and impressions and position become comparable again year on year, subject to any other changes. Quarterly and annual reports that include July to September 2025 will still need the note until those periods drop out of the comparison.

    Hypothetical example:

    An agency’s quarterly report for a client shows impressions down sharply year on year for the third quarter of 2026, and the client asks what went wrong. The agency adds a device split showing desktop impressions accounting for almost all the decline, notes that the drop began in September 2025 with the reporting change, and shows clicks up modestly year on year. It also states that from the fourth quarter onwards, impressions will be compared like for like. The client’s concern shifts to the click trend, which is where it belongs.

    Watch for this:

    The change also affected third-party rank trackers and visibility tools, which altered how they collect results. Year-on-year comparisons from those tools may have their own discontinuities around the same date. Check each vendor’s notes before comparing.

    Other discontinuities to note alongside it

    The September 2025 change is the best known, but it is not the only thing that makes year-on-year Search Console comparisons tricky. AI Mode data has been counted within web search totals since June 2025, without a separate click breakdown. Core updates in March and May 2026 moved rankings for many sites. Consent banner changes, analytics migrations and site releases can shift analytics data independently of Search Console. For each report, keep a short list of known discontinuities in the period being compared, and mark them on charts. Clients and leadership tend to accept a noisy comparison when the noise is explained in advance; they lose confidence when explanations appear only after they ask why a number moved.

    What to tell clients who ask for “the real number”

    Some stakeholders want a single corrected figure for impressions growth. Give them the device-segmented view and a clearly labelled estimate, but be honest that no exact correction exists: Google has not published how much of any site’s impressions came from automated deep result fetches before the change. Offering a range with the method stated is more credible than a precise adjustment that suggests more certainty than the data supports. Most clients are satisfied once they see that clicks, conversions and revenue, the numbers they care about most, are comparable across the change.

    Explain it once, clearly

    Clients and leadership do not need a technical explanation every month. Write a short, plain note once, covering what changed, which metrics are affected and how the report handles it, and link to it from the report’s method section. Consistency matters more than detail: pick one approach and apply it to every report until the comparison periods are clean.

    Use clicks for year-on-year comparisons across September 2025, segment or restate impressions, avoid year-on-year average position until both periods are post-change, annotate every affected chart, and note that from October 2026 the comparison problem largely resolves itself.

    Need help with this on your own site?

    Niraj Raut works with businesses in Nepal, Australia, the UK, Europe and the US on technical SEO, ecommerce SEO, local SEO and AI search.

    See SEO services

Add a reply

Replies are open to approved members. Log in or apply to join.