On a typical installation, Bridg syncs with your POS daily. Bridg leverages this data to track customer transaction behavior, perform predictive modeling and identify segments of interest such as “Lapsed”.
Instead of arbitrarily assigning lapsed customers based on a time frame since last visit, Bridg views the problem differently by tracking every customer individually and comparing their behavior against that of every other customer in Bridg’s database.
Take two sample customers:
Matt typically transacts every week
John typically transacts every month
Shouldn’t Matt be identified as lapsed sooner than John when it comes to the length of time since his last transaction? This is where machine learning and predictive models excel. Bridg leverages the data from all other customers that have transacted like Matt and calculates the probability of someone like him coming back after “n” days of inactivity. When the number of days inactive reaches a statistically significant threshold (typically 90% confidence), that customer becomes lapsed.
This approach allows marketers to send lapsed messages to customers before it becomes too late to matter for that customer’s buying pattern; maximizing the message conversion potential.
The above is performed every night, taking into account the new behavioral signals generated by all of your customers that prior day. As customer behaviors change, so do the predictive models, so that a lapsed customer never goes unidentified.