Selecting Customer Value Dashboard Contents
by David M. Raab
DM Review
February, 2006

Last month's column showed how standard financial information can be transformed into customer value metrics to gain important insights into business performance. But there are many ways to present these values, and the methods chosen will greatly affect the utility of the final product. Here are some guidelines for preparing an actual customer value dashboard.

First and foremost, the information should be relevant to the recipient. Otherwise, the dashboard will quickly become part of the background noise that every competent manager quickly learns to ignore in order to get anything useful accomplished. No one has time to dig through lengthy reports, no matter how attractively packaged, on the off chance that they may contain something important.

One component of relevance is personalization: recognizing that different people need different information and presenting something suitable to each. But listing personalization as a requirement simply begs the question of what each recipient's personalized dashboard should include.

The primary criterion for inclusion is whether the recipient can do anything with the information. For lack of a less awkward term, let's call this "actionability." Sometimes it's obvious how to be actionable: if the Web campaign is pulling in responses at a lower-than-expected cost per order; the system should alert the Web marketing manager and suggest expanding the campaign. But in most cases, figuring out who can react to a particular bit of information is not easy. Does a decline in sales call for action from marketing, sales, operations, or someone else? Often, the best that can be done is to present the information to everyone who might reasonably be expected to respond to it and let them to determine for themselves what to do next.

But this could result in presenting nearly everything to everyone: not a viable approach. A second component of relevance therefore is the business value presenting a particular piece of information to a particular individual. This value combines the ability of that person to act on the information with the expected value of their action. A circulation director may have full control over the costs of subscription premiums, but the amount of money involved could be insignificant. Or, at the other extreme, postage costs may be very important but quite beyond the circulation director's control. There is little business value in presenting the circulation director with either piece of the information.

Relevance also requires that the recipient understand why the information being presented. A single value by itself is very hard to interpret. Most information makes more sense when compared with something else-a budget or forecast, a previous period, a similar line of business. But variances are not sufficient either. Is a 20% increase in customer service costs good or bad? It depends on how much the business has grown, whether the product mix has changed, and many other considerations. The dashboard must factor out such elements before calling the variances to a manager's attention. This is critical, because a system that presents too many false alarms will soon be ignored.

Yet even small variances can be significant if they are caused by large variances in particular business segments. Customers from a new source may have an attractive acquisition cost but also low payment rates. If the quantity is small, the change in over-all payment rate won't be noticed unless payments are examined by source—or until business from that source is increased, with disastrous results. So the dashboard system needs to examine the business in detail, even if the dashboard itself only includes summary figures. Then the dashboard system must alert managers of variances at the detail level that are worth a closer look.

Variances can indicate opportunities as well as problems. Helping recipients to understand the true meaning of a variance requires showing its long-term implications. A small change in retention rates may have little impact on current period revenues but foreshadow huge changes in the future. To identify such situations, the dashboard system must embed its metrics within a larger business model. This model will calculate how changes in each metric relate to over-all business performance. The dashboard system must then use these calculations to help prioritize the information it presents to each user.

Next month's column will look additional factors to consider in designing dashboard contents.

*                       *                        * 

Learn More about CLIENTxCLEINT Customer Performance Dashboard