Tuesday, January 8, 2013

Structured decision making

Making decisions that involve multiple possible options and multiple stakeholders is hard. It's hard because it's usually done in an unstructured way, typically involving a never ending cycle of "pros & cons" that seem to get you into analysis paralysis until somebody makes an "executive decision" just because you've run out of time. As a consultant caught in the middle of a number of key decisions, I've come to rely on a process that I've derived from some structured thinking principles to reduce the time needed to make an informed and consensus-building decision.

I've also made the PPT available if you'd like to see the concept in more detail.

At the core of the process is the separation of the thought process into two halves: Creative solutioning and scientific analysis. Around these core activities are some supporting processes aimed at increasing focus and communication. If you feel the 6 steps outlined below are too heavyweight - cut it down to two separate tasks: Creative solutioning without prejudice; and ranking of solutions in a fair and even manner.

Step 1: Decision Context

The first thing to do is state the question you are trying to ask, along with some supporting context. Keep it succinct & focused and try to avoid scope creep. The context is made up of the following information:
  • The question statement: "What technology will we use for our product database?"; "What material should I use for my bathroom floor?". Avoid stating solutions in the question (e.g. "Should I use Tile or Vinyl" for my bathroom floor?") as it will artificially constrain your solutions.
  • The solution goals: Outline some of the positive outcomes or immutables you're looking for as well as some of the things you're not looking for: "The bathroom floor should be water resistant, feel warm to the touch, and be easily cleaned". Again, avoid contaminating your goals with solutions.
A well defined context is critical to ensure you (and everybody involved in the decision) are focused on trying to solve the same problem.

Step 2: Creative Solutioning

In this step, you begin inventing solutions and listing them without prejudice. Concentrate on inventing potential solutions, not the analytic part of determining if the solution is valid or not. Be creative and try new things or draw on your past experiences. Combine solution A and solution B to form a hybrid solution C. Go nuts! Be sure to include all decision stakeholders in this step and make it a collaborative process.

Just don't start listing "pros & cons" (yet)... keep the task focused on solutions that can potentially meet some or all of your goals.

This is my favourite part - it's fun because it allows me to be creative and exercise my imagination. It's one of the reasons why I became a software architect.

Step 3: Collaborative universal criteria creation

When there are multiple parties involved in the decision making process, there will inevitably be different opinions on which solution is the "best". 

This step puts emphasis on moving the focus away from the solutions and onto agreeing upon a set of common criteria or measurements that will be applied to all candidate solutions. Focus on including all stakeholders and decision stakeholders in this step - it's the only place where they can ensure their concerns are taken into account.

The criteria or measurements that are defined should be unambiguous and easily measured. Ideally, all measurements result in a weighted numeric score, so tallying a winner in the next stage is a simple arithmetic exercise.

Note: Step 2 and 3 can be done in reverse order or in parallel. The key point is that they are separate from each other  I usually prefer to do in the order listed here as it feels like I can be more creative if I don't yet have the criteria in my mind.

Step 4: Analysis & Ranking

This is a purely analytic exercise - simply take each solution and measure it against the agreed upon criteria to produce a score for each solution.

This should be the easiest part of the whole exercise and can amount to crossing items off your checklist of "must have" criteria, or measuring the performance of the solutions under certain conditions via proof of concepts.

Step 5: Make the decision

Hopefully one of the solutions have come out ahead after the ranking exercise and is a clear-cut winner. Now it's time to make a decision and there's usually 3 possible outcomes:
  1. There is a clear cut winner and everybody agrees it's the best solution.
  2. There is no clear winner or a tie. You should go back to your criteria (step 3) and add some differentiation factors (or "nice to haves") to the mix and re-run your calculations.
  3. There is a clear winner, but people disagree with the choice. You should go back to your criteria (step 3) and ensure everything is captured. You are not allowed to disagree with the decision "just because"... it needs to be articulated in the form of measurement criteria.
Once you've made the decision, there's an important exercise to follow: Understand why the solution came out on top. Understand why it didn't receive a perfect score. In some cases, a solution that you initially thought would fill all your needs would only actually get you 70% of the way there, so start planning on how you'll fill that other 30% gap.

Step 6: Record and Communicate

It's vital to record the decision as well as the failed solutions for future reference. When you run into difficulty with the chosen solution and people start saying "I think we should go with option B", you'll be able to easily pull together reasoning why you already thought of that.

Finally, it's really important that key decisions are communicated properly and that the reasoning is available for others to review. It helps keep people informed and educated on the pitfalls you've avoided by going with the solution you've chosen.

I hope this is useful!

No comments:

Post a Comment