How to use the Google Search Console Performance report to measure one website change
Use the Google Search Console Performance report to record what changed in clicks, impressions, CTR and average position around one website update. Define the page, filters and observation periods before interpreting the
Use the Google Search Console Performance report to record what changed in clicks, impressions, CTR and average position around one website update. Define the page, filters and observation periods before interpreting the numbers. A before-and-after record describes an association; it cannot isolate the effect of the update from seasonality, changing search demand or other events. Missing report data is an evidence gap, not proof that the page never appeared in search.
Define the single website change and its exact timing
A before-and-after plan begins with one defined change and the calendar date when that change was completed. Without a single change and a single date, later metric differences cannot be linked to one action using the available report data.
Write one sentence that states what was changed. Record the date the change went live on the published page. Note any other site modifications completed in the same calendar week so they can be listed later as competing explanations.
Keep the change log alongside the observations so another person can reconstruct the comparison. How RoadUp works describes the service workflow; it is a separate overview, not evidence that a particular update caused a search-performance change.
- Write a one-sentence description of the change.
- Record the exact date the change was published.
- List any other site changes made within seven days of that date.
Choose the relevant Search Console metrics
The Search results Performance report includes clicks, impressions, CTR and average position, subject to its reporting rules and the selected filters. It measures search performance rather than providing a complete page-indexing diagnosis. An absent row or unavailable metric does not by itself establish that a page is not indexed or has never appeared in search.
Select the metrics that exist for the affected pages. Record which pages or queries will be used for comparison so the same view can be recreated later.
- Open the Performance report in Search Console.
- Confirm that the chosen page or query appears in the table.
- Select clicks and impressions as the primary comparison metrics.
- Note the dimension (page or query) used for grouping.
Set the before and after date ranges
Choose explicit comparison periods with equal lengths and, where practical, the same mix of weekdays. This makes the comparison easier to interpret, but it does not control seasonality or isolate a treatment effect. Keep dates outside the available reporting window marked unavailable, and check whether the latest data is incomplete before comparing it.
Choose two ranges of the same length. Record the start and end dates of each range and the date the comparison view is created.
- Select a before range ending the day before the change date.
- Select an after range starting the day after the change date.
- Make both ranges the same number of days.
- Record the exact start and end dates for each range.
List competing explanations
Search Console data alone cannot prove that one website change caused a metric shift. Seasonality, competitor updates, algorithm changes, unrelated site modifications and external events can affect the same numbers at the same time.
Create a short list of other events or conditions that occurred around the same dates. These items remain on the record as possible alternative explanations.
- List any sales events, holidays or news cycles in the date ranges.
- Note any competitor announcements or price changes you observed.
- Record any additional site changes made in the same week.
- Keep the list factual and short.
Record the raw metric values
Copy the actual clicks, impressions, CTR and average position numbers shown for each date range. Do not add success language, failure language or causal statements. The report aggregates data by property or by page depending on the dimension selected.
- Switch between the two date ranges in the Performance report.
- Copy the metric totals for each range into a simple table.
- Include the dimension used (page or query) next to each value.
- Save the table with the date the record is created.
Produce the measurement record
A measurement record contains the change description, the two date ranges, the selected metrics, the raw values and the competing-explanations list. It does not contain a conclusion that the change caused any observed difference. Date and sign the completed record.
- Fill a short template with the change description and date.
- Add the before and after date ranges with start and end dates.
- Paste the raw metric values for each range.
- Add the competing-explanations list.
- Record the date the measurement record is completed and the name of the person responsible.
Hypothetical worked example
Hypothetical example: A site owner changes the title of one product page on 15 October. The owner selects the Performance report, filters to that page URL, and chooses the metrics clicks and impressions. The before range is 1β14 October and the after range is 16β29 October. The owner records the raw values shown for each range. Competing explanations noted are a seasonal sales event that began on 20 October and a competitor price change announced on 18 October. The record states the change, the two date ranges, the metric values and the competing explanations without claiming that the title change caused any difference.
- Record the change: title text updated on product page.
- Record the change date: 15 October.
- Select metrics: clicks and impressions for that page URL.
- Set before range: 1β14 October.
- Set after range: 16β29 October.
- Copy raw clicks and impressions for each range.
- List competing explanations: seasonal sales event starting 20 October; competitor price change announced 18 October.
- Date and sign the completed record.
An SEO change observation record
Use this record to keep the observation separate from any unsupported causal claim. Fill every field with the actual values for your site.
Change description: One-sentence statement of the single website change measured Decision: If the sentence names more than one change, narrow the scope to one change before continuing
Change completion date: Calendar date the change was published on the live site Decision: If the exact date is unknown, choose the earliest date that can be verified from version history or deployment logs
Before date range: Start date and end date of the before period Decision: Confirm the end date is the day before the change completion date
After date range: Start date and end date of the after period Decision: Confirm the start date is the day after the change completion date and both ranges contain the same number of days
Selected metrics: List the measures available in the selected report view. Decision: Record a missing measure as unavailable and inspect filters, date coverage and reporting limitations; do not turn an absence into a claim that the page never appeared in results.
Report view and filters: Save the property, search type, page or query filter and other active filters. Decision: Recreate exactly that view for both periods; an empty table is a missing observation, not an indexing verdict.
Paired observations: Record each selected metric in two clearly labeled columns, Before and After, alongside the exact date ranges. Decision: Keep the values and units as displayed; do not compare a period total with a daily average or silently treat unavailable data as zero.
Competing explanations: Short list of other events or conditions that could affect the same metrics Decision: If the list is empty, add a note stating that no competing explanations were identified at the time of record creation
Record creation date: Date the measurement record is completed Decision: Use the current calendar date when the record is saved
Person responsible: Name or role of the person completing the record Decision: If multiple people contributed data, list the primary person who assembled the final record
This record documents observed numbers and alternative explanations. It does not state that the website change caused any difference in the metrics.
Alternatives and limitations
If the selected report view contains no usable observations for a period, record that gap and inspect its filters, reporting limits and data availability before deciding whether to use a later comparison window. Diagnose indexing separately with the appropriate tools.
If multiple changes were made on different dates, either narrow the scope to one change or extend the record to show that separation of effects is not possible with the available data.
If the site uses a CMS that rewrites titles automatically, confirm the exact rendered title on the live page before recording the change date.
Limitation: Search Console data can be delayed and may not reflect recent changes for several days.
Limitation: The Performance report is not a complete indexing inventory. Privacy filtering, aggregation rules and data availability can limit what a selected report view shows.
Limitation: The report does not identify the cause of any change in clicks, impressions, CTR or average position.
Limitation: External events, algorithm updates and competitor actions can affect the same metrics at the same time.
Limitation: Equal-length windows alone do not control changes in seasonality or search demand.
Limitation: Requesting indexing or submitting a sitemap does not guarantee that a page will appear in the Performance report within any specific time.
- Check that the page appears in the Performance report before selecting metrics.
- If data is missing, record the gap and the checks performed; do not invent a reason for the absence.
- If multiple changes exist, choose the single change with the clearest completion date.
- Verify the live page title before logging an automated CMS change.
Sources
Originally published by Dev.to WebDev. Aggregated on AIWithGhost for educational purposes β full credit and traffic to the original publisher.