Customer Recalibration Amidst An Economic Crisis
HOW ARE MY CORE CUSTOMER KPIS PERFORMING?
BY CHUCK DENSINGER – COO
As we established in our overview article, there are six customer data questions you should be asking today. The first of these is: “How are my core customer KPIs performing?” Key performance indicators, or KPIs, are just that: the core subset of metrics that tell you how your business is performing. Customer KPIs measure performance in customer terms—reflecting business performance based on customer behaviors and activity.
Of course, revenue and profit are our main business objectives. But as our colleague Mark Gonzales has pointed out, revenue is not a KPIhttps://www.marketingprofs.com/articles/2017/32037/revenue-is-not-a-kpi-but-these-six-measures-are because it is an outcome, or aggregate, of customer activity; and profit is the result of both customer and operational activity. We measure revenue and profit via our traditional P&Ls, but these need to be supplemented with behavior-oriented customer KPIs. This is because our primary goal is to understand and influence customer behavior to drive business performance—through changes in product, pricing, customer experience, and marketing messages. And now, more than ever, keeping a close eye on key customer behaviors is critical.
The foundational set of customer KPIs that leaders should be tracking is straightforward, and multiplied together they equal revenue:
- Number of Customers
- Transactions Per Customer
- Spend per Transaction
Each of these represents a key behavior or set of behaviors. Every customer counted in Number of Customers has chosen to be our customer, a critical behavior! Transactions per Customer counts the number of times a customer has purchased with us. Finally, Spend per Transaction is, obviously, the average amount spent during each transaction, another key behavior. Importantly, these three KPIs can be driven individually as the objectives of business initiatives (e.g., attracting new customers, getting high-potential customers to visit more often, or getting customers to spend more per visit by adding an additional category or service).
This set of customer KPIs raises a new series of questions. Who is buying and who isn’t? Through which physical and online channels? What are they buying? What time periods are most important, and how are these KPIs changing over time? These questions lead to the groupings, or dimensions, that should be explored:
- Customer Groups (segments)—who?
- Channel and Geography—where?
- Product Category—what?
The first question (“who”) must start with a clear definition of “customer” in your data systemshttps://www.targetmarketingmag.com/article/3-steps-for-brands-handling-customer-identity-management. Next, customers must be grouped into meaningful segments that, collectively, represent all of your revenue, and in which a customer can belong to only one segment. If you have an existing segmentation that is well-established, use that. Or simply define basic “lifecycle” segments—e.g., new, active, best, lapsed, unknown—with simple rules for each that make sense in your business:
- New = new to file within past 90 days
- Best = top 10% of spenders in the past year
- Lapsed = no spend in past 12 months
- Active = all other known customers
- Unknown = revenue not trackable to a known customer (e.g., cash transactions not tied to a loyalty account)
More advanced forms of segmentation will be touched upon in our upcoming articles, but this is a good place to start. The important thing is that all of your revenue is captured by your customer groups.
The “where” and “what” questions—channel, geography, and product category—underpin your P&L reporting today. Your data should be structured so the “who” can be viewed across these dimensions. This will enable you to understand your P&L in customer terms. For example, instead of “sales are down in District 5,” we can ask why sales are down in District 5, and see customer counts, visits, spend per transaction, and which product categories people are and aren’t buying, across our new, best, active, and unknown customers. It’s much more actionable knowing that sales are down in District 5 primarily because active customers aren’t shopping as often, and when they do, they aren’t buying in two key categories. This stops us from wasting time on new customer acquisition, for example, and focus on the root problem.
Time (the “when” dimension) is important for comparing performance between periods of time, including identifying trends and inflection points. As the pandemic plays out, looking for inflection points is probably the most important thing you don’t normally do, but need to now. This involves seeing points in time in your data in which behavior makes a dramatic shift to a new state. Inflection points introduce uncertainty, in that trends are broken and predictions get very difficult to make. The past literally stops explaining the present or predicting the future.
Inflection points can occur at the intersection of any of the KPIs and other dimensions—a specific customer behavior, in a geography or channel, with a specific product category or set of categories. Once identified, inflection points must be pinned in your data, and used to define pre- and post-periods. They will mark significant changes—increases or decreases—in performance. You must be careful not to use pre-inflection point data in your diagnostics, predictions, or goal setting; it has become irrelevant, part of a time that no longer reflects current conditions.
Customer behavior can have a significant impact on profitability, and it is logical to ask if we should measure profit at the customer level. While it may seem straightforward, this concept raises complex questions about the impact of customer decisions vs. business decisions and actions made by your organization. For example, if a customer primarily buys clearance product, are they a good or bad customer? At first blush, they may generate low or negative margin via their purchases. But the reason that product is on clearance is that it didn’t sell at full price. Is that the clearance buyer’s fault? Further, if they don’t buy it, how much more will it cost to liquidate? Did that clearance purchase actually just save you money?
All businesses have examples like these, and it is important to carefully analyze root causes of customer profit variances. Further, studying differences in customer profitability can yield a lot of insight about your business. Accordingly, an additional KPI that can be important to track is:
- Customer Margin Contribution
This is computed differently from one business to another, but the concept is to capture elements of net profit that are attributable to customers, such as product cost, discounts, costs of returns and cancellations, service fulfillment, customer support, ship-to-home costs, and loyalty program benefits. Some organizations attribute direct marketing costs to customers, others consider those company decisions. Allocations such as channel costs, labor, store overhead, paid media, and other operating costs not directly attributable to a customer can be problematic and should be avoided.
Customer Margin Contribution can be very useful, but because of its complexity, you might start with a simpler computation, such as Customer Net Margin, which is simply price paid (net of all discounts) minus product cost. You can add additional components of variable cost later as you mature your data and reporting capabilities.
PUTTING IT TOGETHER
In summary, businesses should be tracking customer business performance through a minimum of 3 KPIs and 4 dimensions, with an optional 4th KPI measuring a component of profitability. Even with this relatively concise set of variables, a lot of insight will result. The who, where, what, and when dimension questions can be answered with details in your data, ensuring that all purchase data contains the customer (if known), channel, location, product, and date-time details needed. Ideally, these KPIs and dimensions should be easily accessible and explorable through a customer KPI dashboard that enables filtering, drilling, and comparisons across these dimensions. Critical inflection points should be flagged. And as you identify important trouble spots or opportunities, you can charge your data science team with probing for deeper patterns for richer insight.
There are many more customer behaviors we can dig into, such as price sensitivity, mobile and online browsing behavior, engagement with services and customer support, media response, and brand and category propensities. But this starter set will help you understand the fundamental shifts underway in your business as we all navigate these unprecedented times.
Footnotes [ + ]