Summary;
- Beyond your top 10-20 accounts, understanding loyalty is really tough
- It requires understanding all of the parts and service transactions with a customer, but in the context of the equipment they own, their operating profile and maintenance strategy
- Historically this has been done via tribal knowledge – that can’t continue, especially with the workforce transition
- AI presents a method to leverage existing data to determine customer loyalty and then take action to build loyalty
In the past, OEMs have depended on account managers and/or dealers/distributors to have a finger on the pulse of their customers. This means surfacing sales opportunities as well as maintaining a sense for how loyal a customer is. This is critical to figure out where to allocate resources, especially scarce resources. An example is where to have executive leaders engage. Of course, they will call on the largest customers. Beyond that, are there mid-sized customers where that executive engagement will actually move the needle? Another example is pricing and discounts – which customers will aggressive pricing actually make a difference in keeping/growing the business, vs others where we are just giving away margin?
As manufacturing businesses have grown and diversified, and the installed base has continued to expand, truly understanding a customer gets harder and harder. Most organizations have put in place some sort of key account management for their largest 10-20 customers where the answer to deep understanding is allocating costly resources in the form of account managers and analysts. For the rest, which often makes up a large portion of the business (and are often higher margin because they do not have as much pricing power), it is much more challenging – with the business either being dependent on the 25+ year tenure sales rep or is completely flying blind.
With tribal knowledge retiring everyday – how to assess customer relationships?
My post a few months ago, “The Great Industrial Resignation” (also called the Great Crew Change) called attention to the challenges that industrial OEMs are facing with regard to workforce transition. Equipment manufacturers are losing their tribal knowledge and customer relationships, with experienced team members retiring in disproportionate numbers (with the pandemic accelerating this). Adding to this is the increased competition for talent that is having an even greater impact on legacy industries such as manufacturing and making it harder to fill open positions. We have also observed a much higher level of employee turnover in many manufacturers, at all levels in the organization, which dramatically impacts performance and necessitates more scalable, process-oriented ways of working
That begs the question – how many hires ‘today’ do we honestly expect to stay for a full career? Or are we fortunate enough to keep a good performer for more than 3-5 years?
Increased complexity combined with the decline in tribal knowledge requires that OEMs think differently about how they gauge their customer relationships. And with the shifting workforce demographic and likely future job tenures, this is going to continue to get more challenging.
Using AI to scale knowledge beyond ‘tribal’
Artificial Intelligence presents an opportunity for OEMs to gain a better and much more efficient understanding of a customer relationship. The strength of each individual customer relationship can be determined by applying purpose-built AI models to an OEM’s installed base and all of their customer transactions. By analyzing similar customers’ transactions across the entire installed base, we can recognize patterns of behavior that indicate if a customer is increasing, decreasing or maintaining their business levels with the OEM.
This sounds simple in theory, but very hard to do at scale. While this could be calculated manually, the complexity of thousands of customers with thousands or even millions of pieces of equipment and millions or hundreds of millions of parts and service transactions make that a massive effort. And of course, this effort would need to be refreshed constantly as new transactions occur. For most situations, manual analysis is not viable because of cost-effectiveness, time-intensiveness or lack of data-science resources.
Purpose-built Customer Loyalty Manager
Entytle has built an AI Customer Loyalty Manager which uses a purpose-built AI model to segment customers based on their loyalty – i.e., Needs-Attention, Below-Average, Average, Above-Average, and Healthy. It has been successfully used by multiple customers across a range of industries. “Needs-Attention” customers can be easily identified and then prioritized for extra engagement to recover or strengthen those relationships. Pricing can be more nuanced and discounting only be used when it is going to be helpful in growing the business. “Healthy” customers can be focused on looking for expansion and/or cross-sell opportunities.
A large machine tool OEM used Entytle’s Customer Loyalty Manager to identify which customers were at-risk before they completely drifted away. They were able to look at those “Needs Attention” customers, prioritize those based on size and potential, and then develop engagement plans to get the relationship back on track. This resulted in multiple situations of being able to uncover the underlying drivers beyond that declining relationship – e.g., service level, quality, etc – and then take action to address them.
In another example, a food processing OEM used Entytle’s Customer Loyalty Manager to help drive their overall customer engagement plan. They had each territory manager look at their set of customers, and use the loyalty scores to define an overall territory engagement plan. This plan addressed how they would allocate their time towards thoughtful ‘saving’ their “Needs Attention” customers that had potential, as well as expanding their relationship with healthier customers. This led to better territory planning, as well as more robust discussions between territory owners and their leadership as to where and why resources should be allocated.
Customer loyalty is very well suited to be easily explained and understood – as well, the barrier to acceptance is much lower, as the AI is simply flagging situations for investigation, as opposed to trying to predict a future event and then take prescription action before its actual occurrence. This has led to Entytle’s Customer Loyalty Manager being a great starting point for commercial organizations to begin leveraging advanced analytics, without a massive change management effort to start seeing results.