Skills
  • Development Direction
  • Project Management
  • Loyalty Program Management
  • Lead Generation

A Loyalty Program Built on a Custom Algorithm Renews Directory

Our contractor directory was a static list of installers and servicers but it was getting outdated or inaccurate overtime. We created a unique in-house algorithm to not only keep the directory up to date, but also rank them based on on a number of criteria to give the end user a better experience.

Companion Explainer Video Made for the Program Debut

Problem

The directory of installers and servicers on our website was highly inaccurate and outdated. Much like an old phone book, the directory was static and had no method of “self regulation”.

  • Oftentimes we would have endusers call us to notify that our listings lead to out of business installers or installers that no longer use our product
  • The integrity of the listings were also called into questions when certain members of the sales staff would “call in a favor” to have a particular installer or servicer to be listed at the top. Eventually over time this would cause conflict as we would lose track of the “favored” listings as the whole directory integrity was already compromised . How do we know who should be first? Who will tell us if that “favored” listing no longer installs our product or goes out of business?

Solution

I devised a new “dynamic” directory that would be based on an algorithm taking into account several criteria. These criteria would allow for a “self regulation” of the directory and would make sure that the listings would provide endusers with the most qualified and active installers in their area.

  • The main criteria was registering serial numbers on the installations. An installer that had more serial registrations would signify that they are a capable installer and also an active installer. This main criteria would keep the directory “fresh” and have all the most active individuals “float” to the top.
  • Other criteria were added to create tie breakers in more competitive markets and also incentivize more loyal and engaged plumbers. These criteria included:
    • The number of our training courses taken
    • The average rating of reviews from end users
    • The last time a user logged into the dashboard
  • For final rank at a local level, the closest 5 within the searched zipcode radius would be awarded temporary “radius points” to serve up the closets, high-ranked companies, within the endusers’ search radius.

Key Takeaways

By making installation registrations a criteria for ranking it helped us “close the loop” on installations happening in the field a bit better. Being a B2B2C company it was often difficult for us to track when installs would happen and where they were happening. Incentivizing the installer helped us with this and gave us meaningful tracking.

Based off these datapoints we were able to “scrape” our database and create a glanceable dashboard to see our top servicers, installers, top products installed (through registrations) as well as the locations that they were being registered among other things

Our algorithm proved to be fair and accurate. We disallowed anyone asking to boost any particular installer but instead taught them all the criteria of the algorithm for them to be able to earn it by their own merit. This allowed for our directory listing to be a source of truth and even had the interesting and amusing byproduct of helping us locate up and coming installers based off of registrations.