Performance Measurement: From Eye Candy to Brain Candy

Our previous issue demonstrated two tips you can use to improve your performance measurement presentations. This issue continues with two final tips (3 and 4) that you can use to make sure the visuals supporting your reporting help your performance measurement system live up to its potential for your organization.

  1. Missing Perspective. When developing your performance reporting, it’s important to understand how perspectives can combine to create the most meaning. If you provide too few, you’ll lose important context, but have too many, and you’ll create confusion and run the risk of “analysis paralysis.”

First, let’s look at some common gaps in perspective. Then we’ll consider some newer methods you can use to provide large amounts of context within a small space.

As companies began to realize how much value was potentially buried in their data, they began to look for more sophisticated packages to help them with their analysis and preparing attractive reports. Vendors proliferated, offering ever more bells and whistles, but not always useful ones.

For example, some carried their dashboard analogy too far, using slick, but distracting, gauges glistening on the page. Whether or not you liked the look, many of these indicators were not only visually distracting, they provided very little context for information while gobbling up large chunks of valuable reporting space. On the right, the gauge is capable only of displaying a year-to-date unit number with a color indicator. What does that tell you?

If you happen to be that one out of every ten men who are color-blind, this presentation is even less useful, since you won’t be able to see the green needle indicating the measured result is a good thing. So while this visual takes far more space than a data table, it provides far less information, and no context for intuiting what you might do next.

    • Are you barely or far above your goal?
    • Is this the beginning of a trend, one past its peak, or perhaps even one entering a steep decline?
    • How well are you doing in comparison to previous periods?
    • Are there contextual elements you should be aware of, such as the impact of interest rate cuts?

If you’re trying to manage from this cockpit display, you’re largely flying blind.

What are some alternatives? Let’s explore some newer formats for viewing a lot of context in a small space. The first is extremely useful when you need to monitor for performance within a specific range. The second is very useful for comparing results to expectations.

    • Monitoring Performance. If the amount of variation within the measure itself is what’s important, Tufte‘s compact sparklines can provide an incredible amount of context in a small amount of space. For example, the sparklines to the right intuitively provide a physician with crucial context for treating his patient by showing recent results with the context of history, and highlighting risky, out-of-range measurements.
    • Performance versus Expectations. When comparison to current period and overall targets is important, Few‘s bullet graphs can be very useful. It may take a one- or two-minute training session to learn how to read them. But once learned, bullet graphs easily compare actual performance to targeted ranges in a compact space.

The examples above focus on creating perspective for a single element. Sometimes the best way to provide perspective is to show how multiple related elements are, or are not, moving in tandem within the same visual. This approach is particularly useful for multi-step processes such as sales or production. Understanding how deals are moving through your pipeline from leads to sales, for example, or seeing how close rates, deal size, and profitability are changing over time can provide far more valuable insight when you show the related factors together. Approaches such as this, for example, can indicate when increasing deal sizes are placing undue pressure on profit margins, or pressure to close deals is having a negative impact on average sale size and profits.

Creating a relational context is just as important for your whole report as it is for each of the individual elements. How have other organizations provided this context?

    • For its silicon technology research, Dow Corning created a dashboard focused on four core elements: revenue, profit margin, customer satisfaction, and one last element crucial to the success of this new unit: an innovation index, which helps assure the company doesn’t sacrifice long-term growth to meet short-term pressures.
    • ABN Amro Bank adapted its reporting to how the information is best used. For example, it uses a push-style performance management model to alert everyone who needs to know when the network is experiencing problems that, if unaddressed, could interfere with its critical trading systems. This allows for activity monitoring, and for resources to be applied to highest priority activities.
    • A major telecom firm, having set the goal of minimizing customer churn, realized through applying context that nearly 80% of the customers they were trying to retain weren’t profitable. The context they had built into their performance reporting system helped them change the goal to reducing churn among profitable customers.
    • In the mid-1990’s, Continental Airlines decided to focus its performance improvement efforts on lowering fuel costs. Unfortunately for Continental, they forgot to consider context, so they didn’t specify any limitations on how costs could be lowered. Soon on-time flights plummeted as pilots slowed their speeds to reach fuel economy targets. (Another airline recently was accused of these same practices.) Customers bothered by the flight delays experienced even more heated dissatisfaction as motivated pilots turned down the air conditioning to save fuel too. Continental’s reporting system didn’t provide context or offsetting goals that considered what its team members might do when striving to reach targets.

So consider shifting your conversations from analyzing history to defining your next steps, by designing your presentation to answer the question at hand in a meaningful context.

Rule of Thumb #3: For insightful performance measurement, first make sure you truly understand the question you’re asking. When you’re clear about the question, it’s much easier to see which elements of context will guide you to understanding and action.

  1. Using Your Best Style. Bar graph? Line graph? Dot plot? Tree map? Sparklines? Bullet graphs? So many styles to choose from! It can feel like facing the red carpet when all you have is last year’s Vogue.Well, not to worry. Your performance reports will always have great style if you take a moment to ask yourself a few questions about what you’re trying to present.
If You’re Presenting… Consider a…
Trends, patterns or exceptions over time Line Chart
Tracking multiple related values over time Bar Graph, or Point Distribution for seeing data clusters
How individual values change over time Bar Graph or Point Distribution
The unique performance or impact of individual items within a trend or pattern, such as months, regions or products Bar Graph
Rankings or rank order Bar Graph
Individual elements requiring focus on their values or supporting comparisons between them Line Chart or Bar Graph
Differences in values that may be difficult to discern Point Distribution and/or a Data Table
Trends over time for a large number of items where the pattern for each item is important Horizontally, vertically, or boxed series group of Line Charts or Bar Graphs (see bar graph and line chart selection guidance above)
An overview of the general relative share of four or fewer elements, when precise values or information regarding change over time is not relevant. (e.g., customer composition or share of top ten customers) Pie Chart Note: Due to difficulty perceiving angles and differences in volume, pie charts should be used sparingly, and with great caution
Concentration or frequency (rather than trend over time) Scattergram or Point Distribution

 

When reviewing your data, it’s likely you’ll find some is better presented as tables, some as graphs, and some as both. For example, in the monthly temperature trends below, the graph provides an intuitive sense of variability and relative warmth in each of the locations. The data table provides a more precise guide, useful for tasks such as trip planning or determining ideal planting times.

Rule of Thumb #4: The style and combination of visuals you choose will depend upon the nature or relationship of elements you’re trying to show. Matching your presentation to the need will help you choose the right graphs or data table formats.

Refining your strategy down to the few elements that really matter requires a commitment to deep thinking and a willingness to challenge your strategy. Choosing performance measurements that effectively determine your progress and lead you to next steps requires deep, analytical thought.

When you’ve invested this much in your performance measurement system, make sure you get more than eye candy. Go for brain candy by choosing your presentation just as carefully as you chose your measurements and strategies. With an appropriate presentation you’ll move easily from understanding what’s happened, to deciding which options will most improve your results.

As Forrester notes, performance measurement projects are a journey. They always have a beginning; but many executives fail to notice that successful efforts don’t have an end. They evolve as a series of victories, each step offering an opportunity to gain insight and evolve measurements to meet the changing needs of your organization.

May your journey be long, and filled with brain candy.