You’ve been sold on this idea of evidence based management - great.  So, what does that look like in practice?  There are a lot of resources and information available about the theory, but you need a decision on that species protection program/underperforming warehouse/share portfolio by friday!  What do you do (other then call us, which you are most welcome to do!) ?

To answer this kind of query, we published a whitepaper “The Decision Making Process - Roles and Responsibilities” as a practical introduction to the art of decisions, using a standard modern enterprise as an example.  The details are in the paper, but we thought it would be useful to outline the basic process here.

Decision HierarchyAnatomy of a decision

The image to the right shows the ideal contributions from each business unit to the task of converting our data assets to information that management can act on.  In the environment sector, the role of observer may be taken by field officers and database managers.

1. Observe

Although we list this as a basis step, using your observers to monitor your own actions as you implement change allows the business to improve itself through the recording and understanding of feedback loops.

Good corporate and public citizens have been collecting observations for, maybe, 20 years now, but that is not the end of the observation story.  To be useful, observations require a full repository of meta-data, detailing the background information about the actual data and its collection.  Without good meta data, your years of database recordings of client churn or lizard population are just a bunch of numbers.

The key with data collection is simple:

True decision analysis will require using your data in ways that the original database designer never conceived of. 

Only complete and thorough recording of meta data will allow this.  Collecting and recording good meta data at the observation stage will allow the data to be used over and over again, delivering huge returns for little extra outlay.

2) Just stop and think

Ok, so you are armed with solid, clear data.  Now think.  Think about what you are trying to acheive. How will you know when you’ve acheived it?  What are the specific indicators of your success?

Many pracitioners would advise that the stage of thinking and planning would come after analysis, but we argue that you cannot undertake an appropriate analysis without defining our terms of reference.

The key here is

The act of thinking is not to find the answer but but to identify what an answer might look like.

Typically the leadership team will drive the thinking process, so that the decision (and subsequent analysis) is driven by their strategic overview.  It is the process of analysis that will bring together the high level strategic needs and the base level data.

3) Analyse

Now we know what we are asking, and we have the data to answer the question, we can analyse.  Analysis is a guided process of transformation of the observed data, through amalgamation and interpretation, into information pertinent to the actions that must be taken.

Your analyses must be tailored to the specific questions asked.  They may require a dedicated team of people and software tools or techniques.

The analysis team should be given flexibility within the scope of the required questions.  It may be, though, that un unexpected analysis outcome results in the action to go back to our observations and ask more questions, so it is quite common to iterate through the bottom three rungs of the process before a final decision is made.

4) Act

It is important to keep your expectations of the analysis phase under check.  The analysis phase will answer your question (provided you have access to reasonable data and good tools and people) but it will only support your decision, not make it for you.

Consider the following possible outcomes of analysis

• There is a significant correlation between the drop in reproduction rates of the thorny-backsided lizard and mining activity on thorny-backsided ridge
• Dry season rainfall in the Wimmera-Mallee shown a significant downward trend over the last 50 years, but wet (growing) season rain shows a flat trend.
• Warehouse X will increase net throughput with increased staff levels, but increase machine maintenence costs.

The final action is predicated upon the information received from the analysis phase, but ultimately a decision is a human endeavor.  A farmer may choose not to modify his farming practices based on the analysis above, because his growing season is not affected.  Another farmer in the same situation may decide to change practices, because he recognises that low off-season rain will affect water table levels in the long run.

Both may be valid decisions. 

Essentially, employing evidence based management techniques can be complicated in practice and involves the interaction of a number of groups.  Decision Analysts work with these groups to streamline and support the process (more about that in the white paper)