Why the name ‘MixShift’?

The name MixShift relates to our ‘Bridge Engine’; one of the most unique aspects of our solution and one which is core to how we think about management. “If you can’t measure it, you can’t improve it” and the ‘Bridge’ was built to programmatically explain changes in performance using mathematical algorithms built specifically for the Amazon data set.

A change in performance can be explained by looking at two components: a change in mix, and a change in rate. Once you account for both factors, you can fully attribute the change in performance back to the component parts.

If I lost you there, here’s an example. If your ACOS (Advertising Cost of Sale) increased from 20% to 25% month over month, we would say that it increased by +5 Points, or +500 basis points (bps). We’ll keep it simple and say that your task is to explain which campaign(s) drove that increase.

The first thing you may do would be to rank your campaigns, from high to low, by their month-over-month increase in ACoS. This is what we would call a ‘rate shift’, and at first pass you may conclude that the worst performer there was the reason your ACOS increased. However, if that campaign only accounted for a small percentage of your overall spend, that increase in ACOS really didn’t have much effect on changing your overall performance. The missing piece is the proportionality, or mix, of your campaigns. Higher spending campaigns will have outsized effect on performance should their ACoS go up or go down. And similarly, changes in spend (mix), with no change in ACoS (rate) can still lead to significant changes in your overall performance.

So, to get to a definitive answer to the question “which campaign(s) drove the +500bps increase in ACoS?”, you would need to account for the mix and mix shift, rate and rate shift, for each campaign. From there, calculate and combine the results to determine the extent to which each individual campaign helped, or hurt, your overall performance.

If that’s too much math, don’t fret, this complex calculation is what the ‘Bridge Engine’ does for you. We calculate the ‘contribution to change’ each Campaign had in driving the change you are seeing in performance (and on down to ASINs and Search Terms for good measure).

And we don’t stop with ACoS. We will ‘bridge’ every metric possible to give you the clearest picture of your performance. Why? Because we want you to be able to trace the performance back to its root cause and put you in position to take quick and effective corrective action, or explain the results in manner that ties back to the overarching objectives.

To that point, a typical talk track could go something like this: “ACOS went up, driven by campaign ‘X’ due to increased CPCs as a result of our customer acquisition initiatives, which increased Sales but at a cost of +500bps to our overall ACOS.”

To bring it back to the question… At MixShift, we aim to provide definitive assessments of performance without the normal squishiness that comes with explaining complex data sets. It’s what makes us unique, and our product sticky, so we decided to give our engine a nod with our namesake.