Original Post Date: Wednesday, March 16, 2011

In Parametrics is Free, I acknowledged receiving (too late) “you should’ve known to ask that” over the years. Quality control after-the-fact is fine; but it’s better and cheaper to take a systematic approach to quality assurance as part of your estimating process. The sheer volume of what we model can often keep us so close to the details that we are unable to step back and put on our QA hat on for a sanity check. Enter Quality! On a very large project, our team has introduced a few regular cross-checks, notwithstanding typical math check-sums.  

  1. A round table peer review, where we describe and defend our parametric modeling approach as well as consistency across similar objects/ systems/ alternatives. 
  2. Introduction of an independent 3rd party, not part of the estimating team, who is an experienced modeler. This resource can take the client perspective in reviewing, say, procurement versus cost-of-ownership as well as comparables from similar program experiences.  
  3. Take an agile approach where your project leader is the “product owner” who reviews all interim deliverables top-down.  
  4. Another option is incorporating a client representative into your pre-release reviews for similar validation testing. 
  5. Finally, we pair off internally to verifying each other’s work against the real-time documentation we keep per common templates.  (Available from PRICE Systems, for both hardware and software modeling inputs). They are great visualization tools for internal parameter-tracking, plus external pre/ post model review with customers.

To summarize QA process, we backward plan enough time from delivery to allow us to bring multiple perspectives and approaches to bear. Quality is still (almost) free; and it’s worth the time… every time! Do you have a best practice quality tip that you'd like to share?  Leave us a comment.