Seasonality vs Trend Separator
Paste daily cost, conversions or clicks and separate the underlying trend from your day-of-week pattern — especially useful for multi-market EU accounts with different holiday calendars.
How to use
Paste daily data
Date,value per line — cost, conversions, clicks, whatever you want to decompose. At least 3 weeks works best.
We separate trend from seasonality
A rolling 7-day average gives the underlying trend; what is left over is the day-of-week pattern.
Read the day-of-week index
See which weekdays run consistently above or below your trend — useful for multi-market EU accounts with different holiday calendars.
Paste at least 14 days of data above.
What this tool does
- Takes daily date/value pairs (cost, conversions, clicks — whatever you paste) and computes a centred 7-day moving average as the underlying trend line.
- Compares each day’s actual value against that trend to build a day-of-week seasonal index — e.g. “Mondays run 12% above trend, Sundays 18% below” — useful for multi-market EU accounts where different countries have different weekly and holiday patterns.
- Needs at least 10 days of data (14+ recommended, 3+ weeks ideal) for the trend and weekday averages to mean anything.
Related tools
- Auction Insights Visualiser — Chart competitor auction data with the same paste-a-CSV approach
- Budget Pacing Calculator — A day-of-week dip can look like overspending if you are not accounting for seasonality
- PMax vs Search Cannibalisation Estimator — Separate a genuine PMax effect from ordinary week-to-week noise
Last updated: 19 August 2026 · Built by the CWA Europe PPC team.
Frequently asked questions
What data should I paste in?
Any daily time series — cost, conversions, clicks, revenue — as date,value pairs, one per line. At least 3 weeks of data gives a more reliable day-of-week read than 1-2 weeks.
How is the trend calculated?
A centred 7-day moving average, which smooths out day-of-week swings so you can see the underlying direction. What is left over after removing the trend is the seasonal (day-of-week) pattern.
Is my data uploaded anywhere?
No. All calculations run in your browser; nothing you paste is sent to a server or stored.