How to measure

15 08 2009

“In god we trust. All others have to bring data.” (W. Edwards Deming)

Starting with measurement is not as easy as it first seems. So you should involve someone with experience. We are very proud that Kobi Vider, one of the world’s most respected professionals regarding advanced statistical approaches and quantitative management, will be acting as our partner. The following article is much inspired from his experience.

Business Goals

Do you know what’s funny about how to measure? Don’t start with the measurements! Start with your business goals. What are the market requirements? What do your customers expect? What does your management require? Uttermost quality? Optimal cost efficiency? Short time to market? Look at your company’s vision statements. Look at the business scorecard. There you will find what your executive management considers to be important for your business. And this in turn will enable to to get senior management attention and involvement.

Information Needs

Break the business goals into information needs. Which kinds of information do you need to find out if you follow your business goals or not. What are the real quality indicators that you need to follow? Defect rates? Or customer satisfaction? Is it neccessary to strictly meet the deadlines? Then focus on milestone trends. Or is it more important to deliver ultimate quality? Then focus on defect rates.

Measure

When you made your information needs clear, then think about measurements. Do not measure what’s easy to measure. Measure what you need to know. Measure output products and measure process performance. Process management tools will help you there. Our Stages process management system contains these kind of measurement features.

Act upon your Measures

Your measurements should be followed by actions. Improve the processes that are out of control or perform out of bounds. If the product quality is not sufficient, analyze why. Missing requirements? Bad design? Poor implementation? And, oh yes, you should have processes before measuring them. Otherwise, you would try to measure chaos.

Added Business Value

Finally, always make clear what’s the added business value of your measurements and the following actions. Let’s say,  you measured too many defects and added peer reviews to your requirement analysis and early design stages. Did the defect rate really drop? How much? What does that mean for your business? In Dollars, Yen or Euro. How much did the customer satisfaction increase? Prove the ROI of your efforts. Your management will ask you about it.

Erich



The different Stages of Process Management

9 08 2009

Almost everyone in business talks about processes as one of the key factors of success. But understanding what needs to be done to actively manage and optimize your processes is not easy. And there is no silver bullet. As always, it depends on your business and what you like to achieve.

Some people think that optimizing processes means drawing some squares and arrows on a whiteboard. If you are lucky, you find someone who draws these pictures in Visio. And – surprise, surprise – for some very agile kinds of business, this might be the best thing to do. IT driven folks might confront you with things like workflow, web services, SOA, BPM, BPEL, BPMN and the all rest. And – surprise, surprise – for some businesses requiring lots of fully automated (and automatable) IT tasks, this might be the best thing to do.

But these are the two extremes, there must be something in between. From our experience, the optimal process management solution for an organisation depends on three factors:

  • business model,
  • business goals and
  • process maturity

If an organisation neither has a business model nor business goals, its chances to survive are not very high. So, we will concentrate on the third one: process maturity. Models like CMMI or SPICE allow organisations to determine their process maturity. The CMMI model focusses on the organisation’s business goals (contained in so-called “policies”) to find the optimized level of process management. As organisations become more mature, they can adopt more and more aspects of process management. Therefore, a process management solution needs to be scalable to cover and organisation’s requirements from the basic level (process definition) via advanced levels (process control) to experienced levels (process optimisation and predictability).

So, a scalable process management solution should provide features for these different stages, namely

  • defining processes
    • modeling processes (simple processes and complex end-to-end processes)
    • managing consistency of processes
    • publishing process descriptions
    • providing role-based access to processes
  • managing processes
    • collecting feedback about processes and managing process changes
    • baselining and releasing process descriptions
    • defining best practice processes as organisational standards
    • modifying and enhancing the process metamodel (defining what are the important pieces of a process)
    • assuring compliance to standards like CMMI, SPICE or ISO
  • controlling processes
    • performing process tailoring to optimize standard processes for specific projects or tasks
    • enacting processes in real-life projects
      • managing process and project documents with access to document and configuration management tools
      • exporting relevant process parts to BPEL workflow engines, development tools like VSTS or JAZZ/TeamConcert, SOA-capable tools, etc.
    • measuring the process performance
    • measuring the impacts of process improvements
  • optimizing processes
    • finding process performance trends over multiple process instantiations
    • optimizing process design (which steps are really neccessary and which steps can be left away)
    • analyzing trends in process performance and managing mitigations when processes are out of control
  • predicting processes
    • finding weaknesses or bottlenecks before they impact business
    • simulating the impact of process changes
    • feeding back process performance predictions into planning

Organisations that head off to this “process improvement journey” with sustainable results might take years to get to the final stage. Always start with the end in mind, but really take your time. Go one step after another. It is nonsense to think about any kind of trend analysis when your processes produce chaotic results because no one really knows the processes and just does what he or she thinks is best.

Needless to say, that a process management tool should support you at every stage, helping you to reach the next one when your organisation is prepared for it.

…and this is one reason, why our process management solution is named “Stages:)

Erich