Which channels are linked to your sales?
Fit a marketing mix model to your spend

A marketing mix model reads your spend by channel against your sales, week by week, and splits the sales into a baseline, a seasonal pattern and what each channel is linked to. Put spend and sales for the same periods in one file and compare channels on the same footing instead of on each platform’s own count.

What your file needs

• One row per week or month, covering at least two years.
• Spend for each channel.
• Sales or conversions for the same period.
• With only a few months of history, or a channel whose spend never changes, the model cannot separate the channels.
• The free tool analyzes up to 10,000 rows; a larger file is randomly sampled to that size.

What you get back

• Each channel’s contribution and share of your sales, with the baseline and seasonal pattern separated out.
• Return per unit of spend and the return on the next unit, for each channel, with ranges.
• Carryover and diminishing-returns curves for each channel, showing where returns start to flatten out.
• Fit checks and a verdict on how far correlated channel spend limits the split.
• The method named and the R code.

A limit worth stating

Exported platform data can show association between spend and results. It cannot by itself prove that one channel caused a sale, and the report says so where it applies. This is the lite version of a marketing mix model: one outcome, spend by channel, and trend and seasonality as the only other controls.

Marketing mix model FAQ