To efficiently and effectively apply the ValueMetrix methodology, you need to ensure that the organization's internal processes
  • Incorporate analyses of the financial Value of Actions as a standard component in plans, budgets, and forecasts
  • Explicitly penetrate the implications of the conclusions from the Value of Actions analyses 
Below follow a few examples showing that analyses of the financial Value of Actions may result in conclusions very different from what management was expecting. The examples have been selected because we have found similar results in several countries and industries, yet they must not be considered as representative for all countries and industries. You may need the help of experienced consultants to determine what is relevant for your organization!

To illustrate this, carefully review the examples below and ask yourself what you - if you were the top manager - would do to master the problems and opportunities indicated by the analyses? If you for example find that high survey scores don't indicate superior performance, just that the customer base is relatively old and not very demanding compared to the scores and requirements you might get from young and well-educated customers in the cities?

And don't forget to check the quality and precision of the data before deciding on which performance improvement actions to go ahead with!

Check the Quality Certificates

A standard feature of the VMx software is a focus on the overall reliability of the analyses where the Quality Certificates show to what extent the data give the full picture or just minor pieces. These quality measures are fundamental as they indicate which performance improvement actions really would make a difference 

Evaluate the strength of the links

All statistical calculations have error margins, also the calculations that show by how a change in the scores for a Behavior Driver would impact on the Desired Behavior. To address this problem, the VMx software calculates confidence margins both for scores and impacts so the strength of the links within the analysis structure are described as correctly as possible, thereby avoiding such traps as saying that a single relationship between two numbers - possibly identified using Artificial Intelligence - verify a cause-and-effect relationship when in real life the numbers just are very unreliable and possibly just the effect of background coincidences

Scores are influenced by demographics

Survey scores are often described as objective truths but be careful: They are influenced by the demographic profile of the respondents! In the exhibit to the left, all the respondents have rated exactly the same experience from an Internet-based bank yet it is obvious that women tend to give higher scores than men and elderly people higher scores than younger ones. Thus, changes in overall scores may actually be the consequence of changes in customer mix!

High scores are not equivalent to high margins

Just because customers give high scores they are not always more profitable for the vendor than other customers. This example is from a consulting firm where the higher satisfaction scores only explain about 1% of the changes in the margins obtained by the consulting firm!

Everything correlates

It is risky and often misleading to assume that a high score on a certain aspect of an experience of a company always shows that this aspect is a major driver of the stakeholders' behavior. These 44 graphs from a retailer survey shows the correlation between different aspects of the customers' experiences of the retailer and the customers' overall rating of this retailer. They all correlate. The inclination of the regression line is more or less the same for all the aspects!

Factor analysis needed to group the significant correlations

The human brain doesn't separate the 44 aspects of the retailer experience above from each other and it is never a single aspect that alone determines what the Next Decision is likely to be. To make proper cause-and-effect analyses and when there are many aspects to analyze, a factor analysis is important as it can group the aspects into statistically significant factors. In the ensuing structural equations, there are typically 3-5 such factors that have a significant impact on the behavior and thus are the behavior drivers to really evaluate!

Different drivers in different segments

When doing the cause-and-effect analyses, it is important to carefully look at the behavior drivers in different segments because they may differ considerably. This example indicates the differences between the Internet bank's active customers and those who now are "non-customers", i.e. they have accounts but neither deposits nor debt. In other cases, expect that elderly customers have other drivers than younger customers, well-educated customers have other drivers than those with basic education only etc. Thus, as stated in Steps 1 and 2, when you select an action you need to define who to influence, what to expect, and how to get the message across to the target segments!

Satisfaction and Recommendations - Emotionally positive but financially questionable

Just because respondents give high scores to a certain Overall Evaluation does not automatically mean that they also are inclined to select the same product or service the next time they need such a product or service. This makes the cause-and-effect analyses critical in connection with financial linkages; the impacts of score changes may be very strong for example relative Recommendations but vary greatly between segments in regard to selecting the same kind of product or service again!