Mix shift
Also called Leading indicator.
The change in what kind of lead a platform buys after the bidding changes. Measured on the attributes leads arrive with rather than on what they eventually became.
Why it matters
It is the only evidence available before deals close, and value bidding does exactly one thing: it makes the platform buy a different mix. Watching the mix is watching the mechanism rather than waiting on the result.
What goes wrong
The outcome comparison is the honest end of the question and it is painfully slow. A realistic ten to fifteen per cent effect needs on the order of a thousand resolved deals a side before a permutation test calls it anything but luck, which for most advertisers is a year. So a working system reads as inconclusive for most of the first year, which is true and useless to somebody deciding in March whether to keep paying. A lead's attributes are known the day it arrives, which is roughly four times the sample with none of the lag, so effects invisible for a year in outcomes are obvious in weeks here.
What it is not is proof the money arrived. It says the machine is doing what it was installed to do. Keeping the two apart matters, because a mix shift with no revenue behind it eventually means the values were wrong, and a tool that let the fast number stand in for the slow one would never find that out.
What this product does about it
100 Google leads either side of the switch before a shift is reported, measured against a fixed model applied identically to both cohorts, and never presented as the outcome comparison.
How to check it on your own account
Compare the share of your leads from your best segment before and after the switch. If nothing moved, the bidding has not changed what it buys, whatever the reported conversion value says.
Next to this
- Learning periodThe stretch after a bid strategy changes while the platform gathers enough outcomes to bid confidently.
- Control groupIn this context, the leads that never came from the channel you changed.
- Attributable liftThe part of an improvement the change can actually be credited with: the gap between how the affected group moved and how the control moved.
- Conversion lagThe time between the click and the outcome.
The rest of this step
Measure this on your own data
The diagnostic reads a CRM export in your browser and reports your volume, your match rate and the spread between your leads against the thresholds in this glossary. Nothing is uploaded and no account is needed.