Trend, season and noise, separated
Time-Series Decomposition
Splits a daily traffic series into its trend, its weekly pattern and what is left, so a real decline stops hiding behind a quiet weekend.
How it works
The series is decomposed into three components: a trend from a centred moving average, a repeating weekly seasonal figure, and the residual. Search traffic has a strong weekday rhythm, so a raw day-over-day comparison is mostly measuring which day it is. Removing the season is what makes the trend readable.
What you get
- The underlying trend line with the weekly pattern removed
- The size and shape of your weekly seasonality, by weekday
- Residuals, where an unexplained day stands out immediately
- A clear read on whether a drop is seasonal, structural or a one-off
What it needs
At least 14 days of daily data, and meaningfully better with 8 weeks or more.
What it will not tell you
It describes what happened; it does not attribute a cause. Annual seasonality needs far more history than Search Console keeps.
Runs on demand
Experimental analyses are not part of the nightly job. You open the report, set the period and run it, so the cost lands only when you want the answer.
The rest of Experimental Analytics
Ready to see what your Search Console data is hiding?
Connect a property and every report fills in immediately. No credit card needed.