The Chief Analytics Officer (CAO): Your New Weapon For Operational Excellence!

Predictive Analysis (PA) can improve your complete business operation.

The Chief Analytics Officer (CAO) can help you with substantial benefits by applying Predictive Analytics (PA). The results of frequent simple and tedious tasks will be faster available, with a higher quality, leading to cheaper operational processes, on the fly innovating your products and services.

Repeating Simple Tasks Over and Again

Did you ever check incoming invoices? Or the expense sheets of your employees? And do you know any colleague that does this same job? And also time after time? Have you ever wondered how much time this takes: Per day? Per Week? Per Month? Per quarter? Per year?

Now I am fully aware this is an important task that cannot be delegated to others. I have done this myself hundreds of times. Many of these items are OK. There is really no need to check these, but you cannot skip them. Only a few are at fault. There are only a few offenders. They make it necessary to check them all tediously.

Would not it be nice if your were presented with only the questionable ones, with the offending reason, with a reasonable accuracy?

The Chief Analytics Officer (CAO)

In any organization there are quite a number of this kind of trivial tasks that need careful attention. The moment you decide to apply some form of process performance measurement by using Key Performance Indicators (KPI's) you are going to spend time on evaluating some item. Whether it concerns incoming invoices, expense sheets or something else, the task adds to the number of tasks you already perform. More work, and more precise, more important work. Because the results immediate influence the efficiency of the operational, tactical, and strategical processes.

A new trend is therefore to have 1 person responsible for analysis and steer information of this type of activities: The Chief Analytics Officer (CAO). The CAO coordinates all analytics activities and makes sure that there is generated a clear synergy where all process benefit from. Important considerations for the CAO activities:

  1. Managing approach, methodologies and tools, and competences for process operation efficiency improvement projects.
  2. Planning and control of such projects.
  3. Linking strategy to performance.

Now you may comment: 'What is new here? I am around now for some time, and I have seen come and go many types of everything solving solutions'. You are right of course, there have been announced many solution types:

  1. Reports (sixties).
  2. Decision Support Systems (DSS) (seventies).
  3. Executive Information Systems (EIS), also called Management Information Systems (MIS) (eighties).
  4. On-Line Analytical Processing (OLAP) (nineties).
  5. Data Ware House (DWH) (2000).
  6. Business Intelligence (BI) (2005).

You miss a few newer approaches? What about these:

  1. Data mining.
  2. Big data analytics.
  3. Business and webanalytics.

Well, I would only be interested in differences in the used methodologies and tools, not the marketing name. Then the analytics list becomes:

  1. Predictive analytics.
  2. Prescriptive analytics.
  3. Descriptive analytics.

Now, what is really different? In predictive analytics (PA) you can predict with a well defined certainty a specific item will fulfill the conditions derived from previous older transactions. An example: You can predict whether the current invoice is OK, or suspect. This is very different from the approach from yesterday. Sometimes, it is called micro-manipulation. Instead of the old approach, aggregating totals and subtotals, or calculating averages, minimum and maximum and other exceptions, your invoice become automatically labeled. Very specific indeed. This opens the approach to replace small tedious tasks with a generic approach that is far superior to to human work: Fast, cheap, and accurate. Rules apply to everyone as they are executed the same way all over again.

Examples of the Usage of Predictive Analytics (PA)

These companies did not miss the Predictive Analytics (PA) boat. I will give some examples of a few of the different methodologies. In most cases the used techniques are the Open Software statistics programming language R, and other free tools like 'awk'.

  • Risk prevention (decision tree)
    The Chase bank used PA for determining which bank loans were not going to be repaid, and remove these in time. Benefits: Saved billions in losses.
  • Detect fraudulent invoices timely (ensemble model)
    The Department of Defense (DoD) used PA for detecting fraudulent incoming invoices. 97% of the faulty invoices is automatically detected. Benefits: Saved billions of dollars.
  • Persuasion marketing (2 level ensemble model)
    Barrack Obama used PA is his re-election campaign to select which voters to address personally, and how. Benefits: He was reelected with a superior majority.

I could give you many more examples, also from my own experience. Feel free to contact me.

Contact Hans Lodder, Chief Analytics Officer.

Spectacular success is always preceded by unspectacular preparation.

Back to top