Levels of Pain – Refining the “Bad” side of the KPI Status

Everything we do towards achieving goals involves costs – sacrifice, investment (a positive way to look at it). We purposefully put things we already have (time, money, our career) to be consumed or at risk in the hope of achieving a goal. For the sake of this short blog, I’ll call that investment “pain”. I present this blog for several reasons:

  1. KPI statuses are usually too simple. Most are a percentage of an actual value towards a target value. But in real life, there are real events that happen at certain thresholds which completely change the relationships between things.
  2. There really are no free lunches. Even if a cost is not readily evident, it will remain a credit in the account of the Complex World to be cashed in sometime, somehow.
  3. Life isn’t really as rigid with artificial boundaries. The systems of life mostly allow for some leeway, even though there may be an optimal value.

With that in mind, I’d like to explore the definition of a KPI’s “bad” status. Measures of a KPI status are not simply on smooth continuum of values (-1 to 1) ranging from terrible to bad to OK to good to great. This simple take on a KPI status was necessary at first because of limited data and processing capability. But in the Big Data era, we can do better by recognizing conditions as a status, not just a calculation.

Events start to happen at thresholds. The bad side of the continuum is really a set of progressively severe stages, each with its own continuum. A nicely graphic example is how the degradation of a car’s tires affect the car’s ability to move along a road progresses through these stages (along with the colors I use):

  1. Warning (pink)- The car tire is balding. This continuum starts with significant wearing of the treads to complete disappearance of it.
  2. Pain (red)- The tire pops, however, a car can still physically move. The tire will progressively shred as the tire rolls along, but you can still get the car off the freeway.
  3. Major Pain (maroon)- The tire breaks off the rim. The car can still move even though it will be sparking on the asphalt.
  4. Broken (black)- The axle breaks. At this point, barring gravity or a tow truck, the car cannot move.

Investment can also demonstrate these stages of pain:

  1. Warning – An investment of money that doesn’t impede cash flow or the ability to keep the “doors open”. The Warning goes away once we’ve recouped our investment.
  2. Pain – We’ve missed payments on a bill. At the high end of the continuum, we begin to receive Dunning letters and even court orders.
  3. Major Pain – Cash flow is impeded, we’re forced to sell of things, our wages are garnisheed.
  4. Broken – The doors are closed on us. We’re out on the street.

Most healthcare procedures such as curing cancer involves trading one Major Pain (death would be “Broken”) for one or more lesser Pains (side-effects).

When taking a 7am flight, but deciding to sleep until 4am, we can:

  1. Warning – Cut it close getting in time to board early.
  2. Pain – Get there boarding late missing out on overhead luggage space.
  3. Major Pain – Needing to run clear across the airport while your name is announced over the PA.
  4. Broken – Missing your flight.

The stages could be different, but the point is that clearly different sets of events are triggered at certain thresholds. Things don’t just become progressively bad from a balding tire to a broken axle. We can live with a balding tire while we ensure our water and electricity stays on.

In the implementation of a KPI capable of reflecting these events, the status calculation requires a CASE-WHEN-THEN statement for each level of pain. To keep with the convention of returning -1 through 1 (bad to good) for a KPI status, I return (-1 and <-.75)=Broken, (>=-.75 and <-.5)=Major Pain, (>=-.5 and <-.25)=Pain, (>=-.25 and <0)=Warning, 0=OK., and anything greater than 0 is Good to Great.

If KPI statuses are set up with these stages in mind, we will have a better idea of the real consequences as we prioritize. Being cognizant of these thresholds helps us prioritize the many pains that we need to address throughout an enterprise or just our lives at any given moment. Prioritization really is deciding what is most important at the moment. But that importance depends on the consequences for failing to address the pain.

If KPIs are the nerves of the “intelligence” of an enterprise, the KPIs should be linked to whatever consequences (effects) that may be triggered. Keep in mind too that there probably isn’t just one progression of pain, but varying progressions depending upon other circumstances. However, that gets into another story, so for now, let’s just start with recognizing that pain isn’t a smooth continuum. (Note: It was this reason that lead me to develop SCL to define relationships in a very robust manner.)

Consider too that warning isn’t necessarily a “bad” thing. As I’ve defined warning in this blog, it very often can be thought of as an investment or a sacrifice that we can live with. As I mentioned, every endeavor towards a goal, the resolution of a problem, requires some sort of investment. In other words, we purposefully put certain things in a level of pain we can live with in order to fix worse pain, which is sacrifice.

This notion of purposefully putting certain aspects of our lives or enterprise into pain (investment, sacrifice) opens the door for a twist on optimization. It is a technique I developed back in 2008 that I call Just Right Targets. The idea is to set up the KPI bad statuses as described here, select a threshold we’re willing to tolerate, select values we can manipulate (in what-if fashion), and run a genetic algorithm finding a set of those values whereby we do not exceed any of the pain thresholds.

For example, the purpose for Project Management is to deliver a product with the features required to provide intended value (scope), on time, and within budget – hopefully without having trashed the “resources” (burning out the workers and wrecking the equipment) for the next project. It’s the project manager’s job to balance the competing goals of scope, time, and resources:

  • If it’s discovered that new features are required, the project manager must negotiate a reduction in a combination of time and resources.
  • If the project deadline must be delivered earlier, say for competitive reasons, scope is reduced and/or resources are added.
  • If the key developer is hit by the proverbial bus, and there is no replacement, need to extend the timeline and/or reduce the scope.

Just Right Targets accepts the maximum acceptable pain bucket for each KPI (at the highest level for this example, scope, time, and resources), runs what-if forecasts on a wide range of scenarios, and scores each scenario based on the minimum pain it will cost, with preference to scenarios that do not cross any pain margins.

About Eugene

Business Intelligence and Predictive Analytics on the Microsoft BI Stack.
This entry was posted in BI Development, Cutting-Edge Business Intelligence, Data Mining and Predictive Analytics, Map Rock and tagged , . Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s