CRO: Customer Revenue Optimization

Customer Revenue Optimization

Optimization has become the buzz term of the day. Let’s optimize our value propositions. Let’s optimize web site conversions. Let’s optimize our media spend. It may seem odd that a piece titled “Customer Revenue Optimization” would start by pointing out that the term “optimization” has come to lost its meaning.

But the reality is that it is the thinking beyond the buzz term that’s lost its way. All too often, firms are focused on optimizing sales and marketing, treating CRM as an expense item to be minimized. While expense reduction is a viable tactic for improving financial performance, there comes a point where pennies saved today will ultimately cost dollars tomorrow.

Moreover, many firms are under the illusion that growth can only come from improved or expanded sales and marketing.

Nothing could be further from the truth.

Every marketer understands the notions of “average order value” (AOV) and “customer lifetime value” (CLV). Sadly, many marketers fail to remember that CLV = AOV x frequency.

Thus, while marketing is lavished with budget dollars to get new customers in the door, CRM often languishes to find the funds necessary to continue engaging those customers and improving frequency.

This makes little sense. Once a customer has been acquired through the marketing function, CRM has an enormous opportunity to improve ROI by continuing to engage that customer and encourage him / her to increase transactional frequency.

In many organizations, sales and marketing are lumped into one category (or at least share some DNA) while CRM is housed somewhere near IT. The reality is that marketing proceeds a sale, and CRM ensures that sales continue well beyond the point of customer acquisition. IT is then the strategic enabler that provides data to both marketing and CRM to improve targeting. This is the model by which the Kabardian Group’s Customer Revenue Optimization (CRO) operates.

Put simply, CRO is about providing more value to more customers and getting more revenue in return…. Without spending any extra dollars.

When done well, from a customer perspective, this means receiving the right offer (at the right price), at the right time. From a firm’s perspective, this means simplifying an otherwise complex marketing program and using data to improve the efficiency of campaigns.

The efficiency gains of CRO lead to less inbox clutter for the customer, and improved sales and profitability for the firm.

Of course, all of that sounds simple in theory, but what about practice? While getting more customer revenue without incurring any additional costs may seem like magic (or a dream), the reality is that a CRO effort can be successfully carried out through a five step process.

  1. Customer segmentation
  2. Value mapping
  3. Messaging matrix development
  4. Offer testing and learning
  5. Operationalization

Customer segmentation

There are countless books written on the topic of customer segmentation. Some advocate segmentation models based on demographics, geography, product affinity or psychographics. Additionally, many firms spend a lot of time and money (often needlessly) attempting to append external data to customer records in an effort to better understand who their customers are.

As human beings, we are often pressed to articulate who we are, making it highly unlikely that compiled data will fare any better.

A far more interesting (and valuable) way to approach the question is by understanding what we do. This is a more objective and measurable means of evaluation, and can be approached by looking at average order value, frequency and product attributes. The first two variables are fairly straight forward and don’t require much more explanation.

Product attributes include any dimension that separates one transaction from another. This might include any of the following:

  • Day / time of purchase
  • Color, make, model
  • Packaging
  • Ancillary offerings that emphasize convenience or cost savings

Our CRO model’s approach to customer segmentation begins by combining average order value and frequency into a single variable and organizing customers into quartile from low to high. The next step is to organize product attributes into four categories: price, value, convenience and premium. Price is a function of frugality. Value is a matter of performance per dollar. Convenience includes any attribute that reduces friction in the purchase, use or return of a product or service, and premium is a fairly self-explanatory description.


Value mapping

At this point, we now have four broad buckets of customers and four product attributes. Depending on the firm’s industry and internal preference, these can be evaluated using sorting algorithms or statistical correlation analyses to segment customers into 4, 8 or 16 different segments.

This is the step in the process where hard and soft sciences meet, and where collaboration between marketing, CRM and IT is necessary. It is a purely technical matter to run an algorithm or statistical correlation analysis. It is far more challenging to develop the assumptions (or hypotheses) used to determine what drives customers to chose a certain behavior.

For example, in our work, we once saw that frugal customers who were focused on price seemed to always select a certain premium feature. This pattern of data caused us to question our entire methodology- until someone in marketing pointed out that the premium feature was being promoted as a loss leader to attract more customers. From this perspective, the pattern made sense. Frugal customers were looking for great deals, and the presence of a premium feature did not contradict the methodology, but rather reinforced it.

The end goal of this step is to combine mathematical analysis with brainstorming to answer the following three questions:

  1. What do we think our customers value?
  2. Are they willing to pay money for it?
  3. How do we know? (Or why do we think that?)


Messaging matrix development

The discussions necessary to address the three questions can often lead to heated exchanges. Data can be contradictory; it can be interpreted in different ways, and personal opinions can cloud matters even further.

All of this is a good thing, though, as these are what will ultimately give rise to hypothesis development and message testing. At this point in the process, our CRO model calls for the development of a messaging matrix. This is a very simple table where the rows and columns represent the customer segments and the value propositions that seem to resonate best.

This simple instrument often helps to resolve disconnects in the data, as it aligns the messaging that generates the most sales revenue with the segments that buy the most. This is a helpful framework for continuing to answer the three questions in the value mapping step in the process.


Offer testing and learning

With the messaging matrix in hand, the next step in the process is to execute a test campaign. At this stage, many will be tempted to look at marketing or financial performance to determine whether the segmentation model is working. This urge must be resisted at all costs. The true metric is not whether a test beat a control, but whether the customer segments seemed to discriminate well against one another.

Put simply, the initial offer test and learn effort should look to prove that customers in each segment behave consistent with others in the same segment and different from customers in other segments. If this holds true, then the segmentation model is on the right track, and financial performance can be achieved by refining messaging , ad copy and creative design in marketing offers.



Although this may seem like a lot, with the right mix of departments, all of the above can be carried out in just a few days. The key is not to get it perfect the first time around, but to begin a process that can be optimized over time. (There’s that darn word again.)

Some might refer to this as developing a data science function. That is often the end result of this process, but the first step is to ensure that CRM and marketing are properly talking to one another. As noted previously, IT is often the key facilitator, making data available (or translating data) as necessary. More often than not, input from finance is helpful, as it ensures that “marketing math” is tied into real financial performance, and that data provided from IT is grounded in the KPIs that are evaluated in different parts of the organization.

Like what you read? Feel free to subscribe to my Lateral Thinking Newsletter. I try to blog once a week, and my entries are delivered once a month to your inbox. Registration is free, and I will never share your email address. You can register here.