Turn guesses into estimates with Wideband Delphi

A group estimation method can yield more meaningful predictions than any one individual can produce

Karl Wiegers
Analyst’s corner

--

Photo of a mystical woman holding her hands above a crystal ball.
Image by Freepik

Most software professionals must provide estimates for their work, but few of us are skillful estimators. Many of us haven’t been trained in estimation techniques. We’re too optimistic, with short memories that mask the painful overruns from previous projects. We don’t incorporate contingency buffers to accommodate unexpected events or risks that materialize. Moreover, we often overlook necessary aspects of an activity. When we eventually encounter those tasks, we either perform them — thereby exceeding our estimates — or skip them, possibly compromising quality in the process.

There are several ways to become a better estimator. The simplest is to record your estimates along with your estimating processes and assumptions, and then record the actual results. Comparing actual outcomes to the estimates — and understanding the differences — will help you do a better job the next time.

Another strategy builds on the principle that multiple heads are better than one. Derived from the Delphi estimation method developed at the Rand Corporation, and enhanced by Barry Boehm, Neil Potter, and Mary Sakry, the Wideband Delphi technique uses a small team of…

--

--

Karl Wiegers
Analyst’s corner

Author of 14 books, mostly on software. PhD in organic chemistry. Guitars, wine, and military history fill the voids. karlwiegers.com and processimpact.com