Friday, 6 August 2010

Developing Decision Making Criteria

Often business rules or guidance are required to instruct or advise people on what to do in given circumstances.  It can be difficult to develop these where there are many considerations or options.

imageTo begin to tackle the problem it might be useful to draw up a decision tree and step through all the questions that might need to be asked.  The example here, whether to retain email as records or not, was drawn using bCisive.  A huge tree results.

dtreeHowever, using a print out of the decision tree, analysis reveals there are possibly only 6 scenarios or situations for which criteria are needed.  Here, the bCisive diagram has been annotated with green and red notes to show the scenarios.  Green indicates a “Yes” outcome – an email should be managed as a record – and red indicates a “No”. 

dcriteria From this analysis it was possible to develop a new bCisive diagram, with the scenarios mapped as “Options”, the outcomes or decisions mapped as “Consequences” and additional criteria or questions mapped as “Requirements”.  Note that it was possible to reduce the 6 scenarios to 4 as the decision criteria for 3 of the scenarios were identical.

image A feature of bCisive is the ability to view the diagram as a text outline.  Copying the outline allows it to be pasted into Word or Excel to create a more traditional document.

image The final product – in this case an Excel table – presents the decision making criteria, outcomes and recommended action in a simple, tabular format.

Although this example has been presented as a desktop exercise using software, the approach could just as easily be followed in a meeting or workshop using flipcharts, PostIts and marker pens.


Sandy Schuman said...

This sounds like a case of "requisite modeling," developing all the model you need, but not more than you need to make a decision. Working with a group, we typically develop a larger model initially than we eventually agree is needed. I would be interested to hear if others have similar experiences.

Steve Rothwell said...

What I also found was an initial reluctance to accept that the decision could be made using such simple guidelines, "surely it's more complex than that - this is too simple". Repeating the exercise in summary helps gain acceptance that it really is that simple. However there are occasions where more detail is added, often spoling the initial result and making the criteria or rules difficult to sell to a wider audience.

Alex said...

While bCisive is obviously designed for this purpose - and as you say, you can achieve the same result using a flipchart and marker pens - most mind mapping programs can be used to provide a resonable software-based alternative.

In MindManager for example, you can start with a simple two-branch orgchart layout representing "yes" or "no". You can represent options by creating an orgchart at the sub-topic level and then switch to a tree layout to show consequences, requirements etc.