In my previous blog (See Business Intelligence, A tip in collecting data for making organisations powerful) I wrote I will give some tips this week on how to collect data for business intelligence purposes.
Organisations that better want to implement, control, change, or delete some internal processes can base their decision on data being collected. For that they usually collect data, they process it to create performance indicators and then act. Those decisions are largely based on data collected and one would like to take logical decisions. However one not always arrives at logical decisions due to the fact that one does not have all the data or correct data.
It is quite important to know if one has the right data and all data needed. But how does one know one has all the correct data? I believe I can't give you a full answer. All I can give is some tips. In this article I want to write something about balancing data.
In balancing data one has to keep the following in mind: The value of data depends very much on how this data can be evaluated. Or one could say data has value to the degree one can compare it with data of similar magnitude.
An example would be that one could compare the monthly income of a company with the monthly expenses. Or one could compare the planned orders this week with realised orders this week. It is a kind of difficult or totally impossible to come to conclusions on the comparisons of amount of registered trademarks this week and the amount of planned orders this week.
One of the ways to find data that is of comparable magnitude is to look at the input, throughput and output of processes. I usually divide a company in 7 main area's of processes. Of these 7 main area's of processes I determine the input, throughput and output. With that I have a first set of 21 main Key Performance Indicators. The 7 area's of processes that I use are finance, distribution, production/services, quality, growth, skills and management.
In finance one can compare the income generated (input), with the goods purchased and paid (throughput) and money on the bank, materials in stock (output)
In distribution one can compare the amount of marketing communications promotions developed and send out (input), with the amount of publications being supplied to bring a customer to understanding (throughput) and the profitable sales being done (output).
In production one can compare the amount of work being planned and prepared (input), with the amount of training to produce being done (throughput) and the amount of production created (output).
In quality one can compare the amounts of quality verifications being done with the amount of quality improvement actions (throughput) and corrections being done afterwards (output).
In growth one can compare the amount of new people being contacted (input) with the amount of people joining the organisation as an employee, supplier, investor or client (throughput) and the amount of people staying or being active for the organisation (output).
In skills one can compare the amount of skilled people hired and people being educated (input) with the amount of communication being done (throughput) and production realised (output).
In management one can compare the amount of idea's, policies, technical knowledge being developed (input) with the amount of registered trademarks, patents, copyrights (throughput) and new plans and new programs being created and executed (output).
From these 7 area's of processes the outputs can be compared.
Within these 7 area's of processes the inputs, throughputs and outputs can be compared.
Of course one can split up the area of production or services. And create multiple streams of production. The same counts for all 7 area's of processes. But basically I think there are these 7 area's.
Within these 7 area's one can for instance split up one process. As an example one can take the process of production and outsource this to another company. This company can then split up this one process of production into 7 area's of processes and define it's own 21 KPI's. You then see that this is recursive. Al 21 KPI's can be split up into another more refined set of 21 KPI's and so on.
Hope this gives another clue to defining KPI's and what data to collect.
Pieter