An annotation is a short, dated note added to a chart in an analytics or reporting tool to record something that happened, such as a website launch, a tracking change or the start of a campaign. When a line on the chart jumps or dips, the annotation tells whoever is reading it what changed on that date.
How annotations work
An annotation holds a date or date range, a short description and, in most tools, the name of whoever added it. It shows as a marker on time-series charts, and hovering over or clicking the marker reveals the note. It never changes the data underneath.
Where you can add them depends on the tool. At the time of writing (October 2026), GA4 lets users with the right access add annotations that appear on report charts. Google Search Console shows Google’s own notices about data issues on its Performance report and lets you add your own notes there too. Matomo has supported annotations on its graphs for years. Looker Studio has no built-in equivalent, so teams often keep a change log as a small data source and plot it on their charts.
A plain shared spreadsheet with columns for date, change, detail and owner works alongside any of these, and it survives a move from one tool to another.
Why it matters
Most of the time spent diagnosing a fall in traffic or sales goes on working out what changed. Without notes, a dip in March turns into a hunt through emails, developer tickets and campaign histories, and the honest answer is often that nobody remembers. With them, the cause is usually sitting on the chart.
UK trading calendars are full of events that move the numbers: bank holidays, school half terms, the January sales, Black Friday, Boxing Day and, for financial services, the end of the tax year. Add your own changes, such as a new cookie banner, a price change or a site migration, and the outside events Google announces, such as a core update, and a year of data becomes readable at a glance.
Annotations also make handovers safer. When a business changes agency, consultant or in-house marketer, they are often the only record of why the history looks the way it does.
Common mistakes
- Annotating campaigns but not tracking changes, even though tracking changes are the ones most likely to create a false trend.
- Writing notes only the author understands, such as “fixed the tag”, with no word on which tag or what was wrong.
- Recording a suspected cause as fact. A note blaming a core update sends the next person the wrong way if the real cause was a stray noindex tag left on after a redesign.
- Adding notes weeks after the event, when the exact date has been forgotten.
- Keeping the only record in a tool you later stop using, so the history disappears with it.
How to act on it
Set up one shared change log now, even if it is a spreadsheet, and agree who adds to it. Record every change that could affect measurement or search visibility: Tag Manager publishes, consent changes, template and URL changes, migrations, price changes, campaign launches and pauses, and outages.
Mirror the important entries in GA4 and Search Console so they sit on the charts people actually look at. When you spot an unexplained movement, check the log first, and use anomaly detection alerts so the next one is caught sooner.
When I run an SEO audit, this history is one of the first things I ask for, because a drop with a known cause is a short conversation and a drop without one can take days to trace.