Working Up

by Dwayne Phillips

I often saw companies blamed for bad management. The real problem was estimating in time of greatest ignorance and customers not mitigating the risk of such estimates.

For a couple of decades, I monitored the performance of companies on government contracts. Part of this job was to grade the companies on how well they performed. The typical grading focused on (1) technical performance and (2) management performance.

The vast majority of the time, we graded companies as good technically, but with poor management. We were wrong.

A typical project performed for the government was based on a proposal submitted by the company. They estimated a schedule and cost by which they would deliver technical products.  Most projects followed this pattern:

• Everyone was happy to start work
• The first six months had a few problems, but we were all still happy
• Problems showed in the second six months – behind schedule and over cost
• Problems continued through the middle half of the contract
• Products were eventually delivered, and people were happy again

Notice the problems – behind schedule and over cost. Notice the happy ending – products delivered. The responsible parties?

• Managers are responsible for cost and schedule
• Engineers are responsible for products

Hence,

• Bad management
• Good technical

I now conclude that these conclusions were wrong. The company was behind a schedule that was estimated in a proposal. The company was over a cost that was estimated in a proposal. The real problem was the estimate contained in the proposal. People didn’t estimate well. And, by the way, most of the estimates were created by the technical people, not their managers,

Why don’t people estimate well? Two factors:

1. They are rushed to make their estimates, often working nights and weekends to complete a proposal in a short time.
2. They make the estimates before work begins, at the point of greatest ignorance.

What do we do now?

Customers, the government contracting agency in this example, need to understand the risk involved in estimates made in ignorance. Estimates are predictions of the future. I don’t know anyone who predicts the future well.

Hence, I advise customers to mitigate the risks with budget and schedule buffers. Add a 25% buffer to your budget and a 25% buffer to your schedule. If you have a million dollars for a contract, adjust the product so that the estimates from companies are about $750 thousand. If you need a product in 12 months,  adjust the product so that the estimates from companies are about nine months.

Customers can scream, “Deliver per your estimate or I will get really mad at you!”

I’ve observed such screaming many times. It has never changed anything.