We use cookies and other similar technologies (Cookies) to enhance your experience and to provide you with relevant content and ads. By using our website, you are agreeing to the use of Cookies. You can change your settings at any time. Cookie Policy.

Driving Business Analytics with the Consideration of Success Criteria

Originally published December 1, 2011

Business analytics programs often center on “decision making.” These decisions might be high-level ones that the top brass of the company are using to steer corporate strategy, or they might be discrete operational decisions made by any number of individuals just trying to get their jobs done. But without the context of a yardstick for success, it is very difficult to objectively say that any decision was a good one.

At the same time, though, a common theme among data analysts and data managers is the concern that the usability of the data within an analytical system is sufficient to enable “good decision making.” That leaves us with a conundrum: the absence of success metrics prevents truly measuring decision quality, but the perception of poor or delayed decisions suggests the absence of success. This Gordian knot can be addressed with a quick slice of the knife by diverting from the implementation of the tools and techniques and considering what is actually meant by “success.”

A review of some of my early articles on straightforward analytics might suggest an idea. If you recall, we have spoken about high-level business drivers and how those can be broken down into more refined aspects of creating or improving value. At the highest level, we can focus on these four areas:

  • Financial, in terms of both increasing revenues or decreasing costs in order to become more profitable;

  • Risk, in which financial, operational, or systemic risks can be reduced;

  • Productivity, leading to increased throughput, decreased need for resources, or increased volumes; and

  • Trust, especially on behalf of the customer and market communities.
That being said, any organization might be presumed to be deploying a business analytics program in order to baseline current performance along some aspect of any of these value drivers and to identify opportunities for improvement. I’ll suggest that this reflection must go beyond the business analytics team, and instead engage the business users to share their experiences and thoughts about what business success means, and then review the existing data to evaluate the situations that reflect the most successful scenarios. The characteristics associated with those success scenarios influence the measures to be used for baselining and targets.

Let’s use a common example in which you want to understand who the company’s best customers are and determine what characteristics they share. Before any attempt at identifying the best customers, though, you have to define what is meant by “best customer.” Although analysts frequently forget to take this step, the definition process establishes the foundation for the analysis. First, the process of specification of “best” essentially provides the yardstick by which success is measured. Next, setting the thresholds for these measures helps in both acquiring the baseline measurement and setting targets for success.

We can take an iterative approach by proposing success measures based on the feedback and input from the business users. To continue the example, I am going to suggest a handful of measures that might contribute to an assessment of best customer:
  • Most profitable over customer lifetime
  • Most profitable over previous calendar year
  • Most profitable over previous 12-month period
  • Longest duration of customer relationship
  • Longest extension of duration of relationship beyond the average customer lifetime
  • Makes frequent orders
  • Makes consistent orders
  • Makes high volume orders in number of items ordered
  • Makes high volume orders in total sales amount
  • Lowest cost to retain
  • Most referrals of new customers
  • Most rapid conversion from prospect to customer
  • Fewest number of calls to customer support
  • Most influential among sphere of influence
  • Most vocal supporter across social network
  • Most reliable as a customer reference
  • Purchases highest-priced products
The value of any of these proposed measures might be refined based on some qualification of how meaningful the measure is to the business, the maturity of the business, and the business’s operating models. A new company might value new customer acquisition, so new customer referrals may be a key measure. A company with products with a long sales cycle might highly value time-to-closure. Some businesses might take a combination of these measures adjusted by different weights at different times of the year.

In turn, that qualification helps to refine the measure. If we are looking for the customers with the lowest retention costs, there must be a way to define what constitutes retention costs as well as ways to measure those retention costs. If we are looking for longest duration of customer relationship, how do we define when that relationship begins? Is it when the individual is initially engaged or when she signs the contract? You can ask similar questions for any of the proposed measures, and eventually you will settle on the criteria to be used to identify the best customers.

The same approach can be used for identifying any success scenario for any aspect of the value drivers, and there are three key benefits that this process provides:
  1. Refinement of success measures based on real perceptions of value;

  2. Framework for differentiation between those scenarios observed to be successful and those that are less so; and

  3. Alignment between business and IT as a result of the interaction in clearly defining the measures.
The next step is to apply those measures for differentiation, perhaps classification, and then analysis of the similarities and differences that might be indicators of success, and I’ll examine that in my next article.

Recent articles by David Loshin



Want to post a comment? Login or become a member today!

Be the first to comment!