Something I've been thinking about for a few months has been how to measure the success of a change in your customer loyalty.
Most loyalty or customer retention apps would like to take credit for any additional sales, but in reality some of those would have happened anyways.
You'd need a large volume of repeat orders to be able to use statistics to prove the benefits (e.g. A/B tests, holdout tests). Even then, those would be less effective for changes lower down the funnel where the volume drops off.
It's almost like you have to fly blind on individual changes and only use wider metrics to guide your thinking. e.g. did Repeat Purchase Rate improve over last year? Is Average Order Value holding steady? This sort off thing my app compares for you automatically as part of its Insights System.
That makes it less about attribution and more about metric monitoring. In a world of imperfect data and rapid change, that might be a good thing.
Eric Davis
Discover where your best customers come from
Going beyond simple attribution, Repeat Customer Insights lets you analyze and segment your customers by who first sent that customer your way.
This will let you find the best sources of long-term customers, not just anyone who orders.