RESEARCH & RESOURCES

The "Key" to Successful KPIs

Sharply focused KPIs will provide the best feedback for your enterprise. Here's how to get started.

By Heine Krog Iversen, CEO, TimeXtender

As business intelligence (BI) practitioners, most of us have a million ideas and just as many activities to complete every day. This volume of information and tasks can definitely overload us.

Enter key performance indicators (KPIs). The value of KPIs is that they help us stay focused at the mission at hand. What happens when we have so many KPIs that we start to lose focus? After all, with a BI system, there are so many metrics that you could measure.

Bernard Marr's Key Performance Indicators: The 75 measures every manager needs to know confounds me. How can a team possibly handle 75 parameters? The answer is quite simple: if you have 75 KPIs, you don't have any "key" ones.

How many KPIs should you focus on? That, of course, is a professional preference based on many variables. However, some people excel at managing multiple tasks, many people focus best on a job task one at a time or, in this particular case, on one KPI at a time. This singular focus helps business executives make decisions based on priorities. I find that being selective can be achieved much more easily when you interact with colleagues from their team, as well as the users of the system, to determine which options to eliminate or reschedule. This collaboration has proven to be beneficial in allowing staff to focus, and with KPIs, it comes down to measuring what matters most (thus the word "key") and eliminating what does not.

Of course, the next question is: What matters most? That, too, depends. For instance, if we're talking about BI for a functional department such as finance, the KPIs in this department will be entirely different than the KPIs used by the marketing department. The KPIs might change from one month to the next based on season or on current need. The point here is that we will need to have a flexible and robust program that allows us to add new KPIs, and to roll out new and better dashboards and reports based on them. This is one of the strengths of data warehouse automation. It enables users to quickly add new KPIs for new interests and areas they need to manage and monitor.

The number of categories for KPIs is quite large. For instance, there are KPIs to support strategy and there are KPIs to support day-to-day operations. KPIs can be considered leading or lagging, inward- (company) or outward-focused (world).

To focus, let's consider one point. When choosing or building a KPI structure for your organization, you will likely need to answer this question: Which set of KPIs will yield the most value at any given point in time?

Even with that, the fact remains that there are just too many KPIs to choose from. How can you decide? Let's examine upward and downward KPIs to consider how they impact KPIs and focus.

Downward and Upward KPIs

First, I suggest you define an evaluation parameter. Whatever goal you want to measure, KPIs tend to work either in a upward (floating) or downward (spiral) direction.

An example of a downward or spiral KPI is cost and profit margin. These elements can only be optimized to a limit. There must be a boundary to how much margin you can generate before you're undercut by competitors due to a price point that is too high. There must be a limit to your cost reduction before you will negatively impact the quality of other services such as customer service or delivery. Once our KPI has reached it optimum value, we will not be able to improve further, and any attempt to do so will require us to micromanage side effects.

However, sometimes, executives tend to take a defensive posture, trying to defend their optimum value within their department, without considering the impact this might have on another department that requires attention and measurement. The super optimization in one department may, in fact, be bad for the entire business as a whole. As a result, it may become harder to reach other goals across the corporation, thereby impacting and hindering a company's overall strategy. KPIs should be designed top down as lower-level KPIs should be designed so they support the upper-level KPIs to secure alignment. Using our example of profit and cost margins, the problem here is that the optimum value may just be unknown. To some people, it may seem that they will always be able to increase margins,, but results may spiral downward, such as reducing costs too far, hence the name "spiral downward."

Another example of a downward KPI system is just-in-time (JIT). Once you have optimized your supply chain to become JIT, what will your next improvement be? "Even-Juster-in-Time"? Again, once a limit has been reached, it's easy to again go on the defensive. Attempts to further improve supply gets pushed out further, putting pressure on other supply chain partners. This could possibly lead to the detriment of the whole JIT effort.

A downward KPI has a built-in,unknown limit. Optimizing this KPI beyond its optimum tends to generate other variables, which then need to be micromanaged; thereby wasting time, money, and focus.

Why, then, are so many downward KPIs, which we know are self-restricting, being used in decision support? The answer is that working with downward KPIs tends to be easy. They are simple, easy-to-understand, and easy to find. This is why decision support often revolves round them and has done so since BI became pervasive. As it turns out, complacency, laziness, and human nature tend to creep into our BI work and impact the efficiency of the platform, and the goals set by management when it decided to invest in BI.

Now, let's consider upward (also known as floating) KPIs. There are quite a few KPIs that have an opposite effect. The more you focus on these KPIs, the higher your return. The potential to drive value is limited only by time and the attention we dare give the KPIs. They offer the ability to draw in more stakeholders. They tend to unite. KPIs that focus on collaboration, innovation, and other cross-functional behaviors have these qualities. Although upward KPIs tend to require more work, the results of BI when the program embraces upward KPIs is rewarding.

Data Accumulation

The other part of this analysis, which is absolutely "key" to consider so we can focus on our KIPs, is what happens when your BI system accumulates a volume of information over time? Some of the data will be bad information or will become irrelevant. The more legacy data we accumulate, the harder it will become for users to focus. At this point, it will be time to prune and eliminate legacy data.

Many of us find it hard to discard personal items we have gathered over time, and when it comes to information collected from a BI system, it can be just as hard. Even more, as IT storage is inexpensive, why even bother with the exercise of recycling and trashing? The answer of course is focus. Too much information is distracting and takes away with what we set at the beginning of setting KPIs, so it is imperative that recycling and trashing be part of your BI program.

To recycle, select what is of value and how it can be re-used. Understand how facts and figures relate to each other and why they were developed in the first place so that the business is working towards unified goals. It's also important to make sure that the KPIs are well documented so that executives can later remember why they were established. Having the knowledge to understand how a KPI came to life, how it has evolved, and who is using it will enable you to separate good information from the bad information. It is essential to have a BI environment where everything gets automatically documented; having documentation is vital and we likely will not have the time to create it manually as we begin to recycle.

Regarding trashing, it is the role of the business leaders to understand which elements are worthy of the BI team's attention. The team can reduce the stockpile of KPIs and help to accelerate the process of removing excess.

Having a focused KPI program is the key to success. Spending a little more time determining exactly what we want to measure and derive from our BI system is paramount. Not shying away from a bit more work and focusing on upward and downward KPIs will help us achieve focus. Eliminating information that we have gathered over time will aid in building KPIs to provide us with a clear vision and strong focus.

Heine Krog Iversen is the CEO at TimeXtender, a leading data warehouse automation company for Microsoft SQL Server. He can be reached at hki@timextender.com .

TDWI Membership

Get immediate access to training discounts, video library, BI Teams, Skills, Budget Report, and more

Individual, Student, & Team memberships available.