Archives for September 2019

Today’s lesson in how not to design a system comes from the U.S. Department of Education (DoE). The system is the public service loan forgiveness program. Congress directed the DoE to forgive college loans for students who went into public service.

In 2018, the GAO found that DoE rejected 99% of applications to this program. They distributed just $27M out of $700M. About 71% of rejections were due to a technical process issue.

What are horrible design! If 71% of your users can’t follow the process, it’s not your users that are the problem, it’s your system! And, by definition, these were all college-educated users! It almost appears like DoE deliberately created a bad system. Probably not though, when the choice is incompetence or maliciousness, it’s almost always incompetence. To its credit, DoE agreed with GAO’s report and recommendations to improve the program.

Further details in the GAO report here.

A great article on six sigma here. It studies the history of Six Sigma (Deming and Japan), its association with Motorola and GE, a hilarious send-up by the TV show 30 Rock, and the variability of certification.

I remember working with the Six Sigma people at Microsoft. I didn’t understand how it was supposed to work there. Microsoft, like most American tech companies, manufactures nothing. All manufacturing is outsourced. And they produce very manufactured goods outside of the Xbox. It was odd to see Six Sigma applied to order processing and customer service.

You must track time against projects in order to do Schedule Management. Period. There are no alternatives. You must track time.

Most organizations don’t track time. Punching the clock is seen as a blue-collar job requirement, not for us. No! You must track time.

Some organizations think the time tracking is micromanagement. It’s not. The results can be anonymous. You must make decisions based on data. You must know how much time your employees spend on operations so you know how much time is available for projects. On projects you need to know how much time you spend on analysis, on coding, on testing, on training. You must track time.

Some fear that the data will be used to punish employees. Simple fix – don’t do that. Don’t tie your time tracking system to performance reviews or even payroll. Use it solely for increasing efficiency on projects. You must track time.

Some say it’s time-consuming. It’s not! It takes me ten minutes a day to do it. I’ve done it every work day for nearly 25 years. It’s easy. You must track time.

Apple has over $260B in cash reserves. Microsoft has $130B, many companies have billions upon billions in cash. Doing nothing.

This means that Apple, and other companies, have more money than good ideas. They have the money to do virtually anything! And they’re not. The great stagnation, I guess.

It’s ironic but the most important project management decision is made before any project management is done. The most important project management decision is deciding what projects to do.

No one knows, or cares, whether the iPhone was delivered on time and on budget. It has delivered tremendous value to Apple. Conversely, it doesn’t matter if the Microsoft Kin mobile phone was on time and on budget. This $1B project was killed two months after launch.

Make sure the projects your organization is doing support your strategy. Kill the one ones that don’t. Measure the outcomes later to ensure the expected results were achieved.

Every now and then I think about NASA’s budget in comparison to other budgets. NASA’s 2018 budget is $22B; Microsoft’s sales and marketing budget is $18B. Close!