About Our Business Intelligence Solutions

I was once invited to a client site to help “create some key performance indicators (KPIs)” for their business.  The only representatives from the client were two information technology (IT) managers and a young project manager, all of who were mostly focused on IT infrastructure projects, and knew little about the business itself.  No one from the business side of the company was present.  Without their input, the IT managers struggled to figure out what chunks of available data were most important, and spent time guessing as to what executives wanted to see.  We sat looking at each other for quite some time; the discussions were tedious.  

It seemed upper management had ordered “some KPIs” with the expectation that something relevant could simply be ethereally conjured.  With terabytes and terabytes of data available, it simply didn’t make sense to approach this problem from the bottom up.  To solve this problem, the discussion needed to start from the top, with the business, not the IT guys.   

Key performance indicators for your business are just that.  They are indicators driven by your business that leaders use to inform their decisions.  Such indicators are typically part of an executive dashboard or business intelligence (BI) solution.

It is important to note that the “lower” we go in the organization, KPIs tend to be less complex and transformative in nature, and nothing more than basic metrics.  For more information, refer to “Key Performance Indicators: Ideation to Creation” in the March 2018 edition of Engineering Management Review (Bishop, 2018).

Understanding where the data comes from or how to acquire it should not be the initial primary focus. These source and acquisition problems can eventually be solved. But, without input from the business on what is most important to them, the impact of what gets created, will be marginal at best. 

Having been in these kinds of discussions in many different contexts, I came up with five important characteristics that not only ensure you’ll have great key performance indicators, but “killer” indicators. Killer performance indicators are relevant, informative, research-based, actionable, and predictive.

These characteristics are not “steps” to creating them. Refer to Bishop, 2018 for those details. Also, these characteristics are not presented in any special order.  The more these characteristics are integrated into your KPI solutions, the more effective they will be.

1. RELEVANT:  KPIs should be based on business needs first.

It all starts with the business.  Development of good business intelligence is a management problem, not a technical one. 

Any conversation regarding business intelligence development begins with executive management.  What problem(s) are you trying to solve?  What kind of intelligence would make a significant difference in your decision-making?  What questions do you wish you had answers to? 

KPI development is often mismanaged in the same way agile adoption is, being driven from the bottom up by developers or project managers instead of executive leadership and the business from the top down. Strategic business relevancy requires top management vision and insight. A technical expert is not necessarily a business expert; expecting such to evoke relevant KPIs from the depths of large magnitudes of available data, will likely bring lackluster results. 

2. INFORMATIVE: Killer KPIs are informative indicators that tell you something about your business you don’t already know.

At one job long ago, I was asked to take over “phase three” development for the company’s business intelligence solution.  I quickly realized that the contractors working for the past three years had done little more than port data over from several existing systems into one location, along with some nice visuals. It saved the user time by providing one central location to find their data, but provided little else.    

KPIs are part of an overall business intelligence solution that is usually based on an extract, transform, and load (ETL) approach.  All too often the “transform” aspect of ETL is left out, with most KPIs being nothing more than porting data over from some antiquated system into a fancy business intelligence dashboard. 

Killer KPIs consider business needs, analyzes data from multiple sources, and juxtaposes them using research-based algorithms to create NEW KNOWLEDGE or insights that would otherwise be unavailable.  With this approach, executive leadership now has special knowledge they didn’t have before, and perhaps even their competition doesn’t have. 

The KPI solution is transformed from a mere data aggregator, to a truly informative and transformative business tool that provides competitive advantage.

In the example mentioned above, I revamped “phase three” into a whole new BI strategy that followed this ideal.  What resulted put the business at a major competitive advantage compared to their industry peers. The company continues to reap the benefits to this day.

3. RESEARCH-BASED: Composition of killer KPIs should be based on research and engaged scholarship.

Although the needs of the business should drive KPI development, such development should be based on research principles to ensure validity, accuracy, and relevance.  Approaches based on anecdotal evidence, trial and error, or tribal knowledge can be error prone or inaccurate.  This characteristic is particularly important when building complex, transformative KPIs that provide new knowledge based on computation and analysis, rather than simply porting data from one system to another.

Research-based characteristics require good algorithms. An algorithm is a process or series of steps for performing a calculation, processing data, or solving a problem.  For example, let’s say an airline executive wants a KPI that measures the profitability of each trip.  There are many inputs necessary for calculating that profit; fuel, personnel, aircraft maintenance, passengers serviced, and cost of meals, to name a few. The calculation can become even more exact by determining when flight attendants and pilots log in and off the plane, the various types of seats being sold and how much money is made on each.  Something that may seem straightforward, could become complicated to create.  This is a simple example. Algorithms can become even more challenging if they include esoteric and intangible data, such as employee morale, customer satisfaction, or market pressures.  These algorithms need to be based on research, instead of opinions or supposition, to ensure validity.  Most importantly, what use are your KPIs if your decision makers do not trust them, or feel like they have to double-check them before making decisions?

4. ACTIONABLE: Killer KPIs can be broken down into components or “controls.”

A good KPI is not only informative, it is actionable.  A killer key performance indicator can be broken down into components or “gauges and dials” so that users actually have tools to facilitate decision-making and actuate change.  These components could be smaller, more narrowly defined metrics or gauges, but could also be something that could be changed like a dial, such as resource allocation. 

As an example, a KPI may show that a company is losing money; but where or how?  The user needs to be able to “slice and dice” the indicator or break it down in such a way as to identify where the greatest loss is occurring, and preferably, why.  Your executive management should be able to use the KPI to clearly identify the problem, and determine a plan of action without needless queries or verification.

5. PREDICTIVE: Killer KPIs are predictive, allowing leadership to visualize tomorrow, as well as today. 

These days business intelligence has transformed into business analytics.  Such solutions not only provide information on the current state of the business, but also the ability to perform predictive analysis.

High rates of technological change are resulting in increasingly turbulent and dynamic markets.  This disruption has created the need for tools that inform leadership decisions on how to reduce risk, adapt quickly to changing conditions, and above all, garner the greatest market share in quickly evolving markets. 

The good news is that the typical enterprise has more available data to mine than ever before.  The challenge is how to leverage such big data analytics into a comprehensive business intelligence and business analytics solution. 

Unfortunately, the bad news is that our capacity to create data has vastly exceeded our ability to make sense of it. Many companies are still struggling with business sense making.  Following the five characteristics will help you navigate the rough waters of these unknowns; emerging from the other side of the data swamp with a high value business intelligence solution.  Our BI solutions at BlockchainFOB are developed with these ideas in mind.