Sunday, January 2, 2011

As Engineers we make decisions all the time, why do we need a DAR Process?

Dear Appraiser,

Our Lead Appraiser is not satisfied with the way we make engineering decisions.  We've tried to tell him that we hire really good people that are really smart, and making good decisions is part of their job description.  What's wrong with this?  Why do we need some heavy process to manage what we already to really well?

Yes, really great people doing great things. . . .why have any process at all?  Boy, you ARE smart!

Many who read "Ask The CMMI Appraiser" have noted that I don't think much of the "advice" you get from some of their so-called consultants, but in this case I'm going to have to call you out.  Good people don't always make good decisions because there are many other factors that affect those decisions - like being overworked, underpaid, and not having the resources to do a quality job.  In other words, good people are human, and humans are fallible.

You're right about not needing a "heavy" process - none of your processes should be "heavy."  They should be useful.  But it's a false choice your presenting.  It's not "heavy or nothing."  Process lives along a continuum, and DAR is the perfect process area to make this point.

DAR is your friend, and it exists to ensure that your buddies and your management don't trash you and your decision after you make it.  It will reduce the chance that your decision won't cost the company millions in unexpected costs (and thereby putting you on 99 weeks of unemployment), and will ensure that whatever you've decided won't be turned around the minute you move up (or move out).

To illustrate the value of DAR in my classes I ask the students to write down the "3-Ds."  Durable,  Defendable, and Deliberate.

A decision is said to be Durable when it stands the test of time, outlasts the person making it, and stays the right decision long after it is made.

A decision is said to be Defendable when it can survive the inevitable sniping be detractors who are offended that they were not asked to make it, those trying to score a point by proving you wrong, or just boneheads in management that think "adding value" means to question every decision a  subordinate makes whether they understand it or not.

Finally, a decision is said to be Deliberate when (and this is ironic . . . ) a well-defined and executed ENGINEERING process is used to make it - thereby supporting the Durable and Defendable parts...

So that's why DAR is your friend, and a good process to use when you need to make any kind of important decision.

Oh, and by the way, smart is not a differentiator . . . most engineers are smart.  It's obviously not enough to be smart.  And using it as an excuse to not be even smarter, well, that's dumb.


Hillel said...

~ Awesome!

Anonymous said...

Hi Jeff

Someone asked a right question at the right moment :).

My question is in regards to evidence for DAR. For making decisions we plan to use consensus based decision making model and when it fails we go for authoritative decision making model. As you know consensus is a meeting approach where everyone sit together we generate lots of alternatives and come to decision.

Does a meeting record of that meeting a good enough evidence?

When we switch to authoritative model,in which a single authority makes the final decision. What should be an evidence in this case? If he take pictures of whiteboard and save his email communication, will that be enough evidence?

Jeff Dalton said...

Good question. For a consensus meetings it would be good if you capture your alternatives and discussions on a white board or in notes. That way we could see how you made the decision, and maybe how to improve that process in the future.

Not all decisions are subject to structured decision making, so authoritative decision making CAN be OK in many situations, but be sure to identify when that can be accepted and when it cannot.