14 August 2009

Print

Decisions, Decisions—Minimize Bias in Risk Management

By Carl Pritchard, PMP, PMI-RMP

Decisions are a critical outcome of effective risk management.  But often, those decisions are rooted in the “salesmanship” and “numbers-crunching” of individuals with bias.  True risk governance means that we minimize that bias any way we can, and, as a result, get decisions less likely to be driven by external influences.

Examine virtually any turmoil, and you’ll find that its roots are frequently found in poor information.  Whether you examine politics, economics or personal decisions, the heart of failed decisions often comes back to someone who lacked information or lacked good information—or let that information be clouded with bias. 

The reality is that anyone who provides information regarding decision-making creates bias.  It is genuinely ubiquitous.  Whether you’re deciding what time your children should be home on a Friday night, or deciding whether or not to invest corporate funds in a particular proposal, you have a bias. 

Some organizations attempt to scrub away that bias with numbers.  They hope that their ability to develop the perfect return on investment (ROI) formula, or to properly generate business cases rooted in internal rate of return (IRR) will expunge the bias from the process.  Numbers don’t lie, they contend. 

Yes, but numbers also don’t speak the truth.  Numbers are nothing more than sterile bits of information, lacking context.  And as soon as individuals begin offering context, they take on the role of “selling” the numbers.  The vicious cycle begins anew.

Organizationally, the answer is constancy.  The more you can do to provide a measure of consistency in how and where the numbers come from, the more they take on a semblance of truth.  And while you, as project managers, may not have the authority to strip this role from marketing or business development, you can still work to ensure that the correct questions are asked on a project-by-project basis. 

For the relative threat or reward, you can ask the direct, simple questions of:

• Who provided the data?
• How current is it? 
• Was it derived in the same way it’s derived for other efforts of a similar nature?
• Are there other sources for the data? 
• Have they been checked to ensure they say the same thing? 
• Are we calculating the outcomes using the same tools, formulae and metrics as we used on other projects?

These questions provide clarity on what you really do and don’t understand.

For those organizations advanced enough to be investing in Monte Carlo or expected value analysis, the probability considerations also have to be taken into account.  Here, beware the individual who pulls a probability out of thin air. 

Instead, you need to have qualitative terminology to back up the percentage values that might be applied.  There should be terms for 90% (almost certain) and 30% (happened more than once, but not on more than half your projects) and any intervening values you might choose.  There should be clear, objective tests so that you affirm that your vision matches your teams’, ensuring consistency in how this information will evolve. 

And if no one has created the tests for your decision-making processes?  That means you have a wonderful opportunity to generate them yourselves.  It is a grand chance to build your professional authority.  Do it consistently over the long term, and you become thought leaders on risk governance within your organizations. 

That’s a great place to be.

Carl Pritchard, EVP, PMP instructs the e-SeminarsWorldSM course, Managing Multiple Projects. He is president of Pritchard Management Associates, a presentation, training and consulting firm that firmly believes that project management should be fun and memorable. Mr. Pritchard is also a PMI SeminarsWorld^® leader and a chapter author for A Guide the Project Management Body of Knowledge (PMBOK^® Guide)—Fourth Edition. He has taught risk and project management around the world since 1993, and invites your questions or comments.