Matomo Traffic Dropped? Check Tracking Before Blaming SEO

If Matomo suddenly shows fewer visits, first establish whether you have a real audience decline, a tracking change or a reporting problem. Those require different responses. An AI summary can organise the investigation, but it cannot establish the cause from a single downward chart.

1. Check that the comparison is fair

Confirm the website ID, timezone, date range and active segment. Compare completed periods with the same weekday mix where possible. Today is not comparable with a completed yesterday, and a month to date is not comparable with the entire previous month.

Check whether you are looking at visits, pageviews or unique visitors. A change in one does not necessarily imply the same change in another.

2. Check whether collection still works

Visit a page where tracking should run and inspect the tracking request using your browser’s developer tools. Confirm that it reaches the intended Matomo installation and uses the correct website ID. Respect the site’s consent configuration while testing.

Review recent releases: tracking code changes, a new consent manager, tag-manager publications, security headers, redirects or a domain migration. A measurement change can affect the reported audience even when demand has not changed.

Use Matomo’s discrepancy checklist to broaden the investigation. Do not change consent or exclusion settings simply to make the numbers rise.

3. Check reporting freshness

For self-hosted Matomo, investigate archiving failures, scheduled tasks and server errors. Reports can lag behind incoming activity if processing fails. The auto-archiving guide explains the operational setup.

Compare a completed historical date with a recent date. If old reports look normal while recent reports remain incomplete, that is a reason to inspect processing, not proof that collection is healthy.

4. Localise the decline

Break the change down by acquisition channel, page, country and device where those reports are available. Ask whether one source or page accounts for most of the decline, or whether everything changed at the same time.

For a search-only decline, inspect Search Console impressions, clicks and affected URLs. For a campaign decline, inspect campaign activity. For a broad decline coinciding with a deployment, prioritise the tracking checks.

These are investigation paths. Correlation with a release or campaign ending is not by itself a causal conclusion.

5. Use AI to assemble the evidence

Savigny can retrieve visits, channels, pages, referring websites and countries for selected periods. After connecting Matomo, try:

Compare visits for the two completed date ranges I provide, using
the selected website's timezone. Retrieve channels and leading pages
for both. Show which observed changes account for the difference.
List tracking checks separately from possible audience explanations.
Do not claim a cause or treat missing data as zero.

Savigny’s current tools do not inspect your deployed tracking code, consent configuration, device reports or Search Console. Those checks remain separate. Ask the assistant to say what evidence it cannot access.

Record the finding

Write down the first affected date, the reports checked, the relevant release or tracking change and the evidence supporting the conclusion. If you correct a measurement issue, annotate the reporting history so the next monthly review does not treat the break as a marketing event.

Use the monthly report template to preserve the caveat and the next action.