In part 1, we looked in detail at what it means by wallet share leaking and how organizations are leaving a lot of money with high margins on the table by not harnessing their wallet share or entitlement in an institutionalized way. We also touched upon the fact that in the world of AI by mining appropriate data and patterns across the services and sales transactions that are happening across the organization, an organization can control its destiny by surfacing every customer in the system and identifying their purchasing patterns and aligning their sales, services, and aftermarket teams around those insights.
But there is an elephant in the room and that goes by the name of data quality. Bigger organizations suffer from the fact that not only their data is dirty but it sits in silos across many functions. Few long-service employees of the organization understand the breadth and depth of data and even no one individually carries the complete picture.
Garbage In Garbage Out aptly summarizes the state of AI and analytics in many organizations. Resources in terms of time and money are spent on analytics but with suboptimal results and in the worst case insights which are bound to make things worse. The first order of things that an organization should embark on is to put data prep pieces in place and ensure good quality data makes its way to the wallet share leaking model.
Coming back to wallet share, once good quality data is in place the whole process can be summarized in four key steps:
- Find clusters of similar behaving equipment.
- In a given cluster, find the best “behaving” equipment and use them as a benchmark.
- Compare the actual purchase rate of all equipment in the cohort and measure it against a benchmark
- The ratio of actual to benchmark is the wallet share leaking gap or entitlement gap
Let’s look into how we cluster similar-behaving equipment. This needs input from business folks and insight from fields. Equipment that is installed in fairly similar conditions can be clustered just based on their duty hours or hours of operations. However, equipment which are exposed to the environment, dependent on many external factors would need multiple dimensions to cluster them appropriately.
Once we have clustered the equipment appropriately it becomes easier to compare the behavior with each other and find the benchmarks. After identifying the cohorts the next step is to find the purchasing patterns as seen in the past transactions.
In part 3, we will look into detail how to find purchasing patterns benchmark those rates, and use them to find the wallet share gaps.
Gain valuable insights through the life stories of Industry Experts, and subscribe to the Aftermarket Champions podcast.
