By Tom DeMarco

Tom DeMarco was a co-author on Peopleware which is why I decided to try this book.

I'm hilariously bad at explore-exploit problems. I get stuck in local minima all the time because I'm scared of "wasting time" in case exploring or trying new options yields no or negative results. I cook from a small subset of meals since I'm scared I'll waste my time making something I don't like when I could have just exploited the mediocre option I know. This doesn't make sense in a holistic view but when your living groundhog day, if you don't want to risk a bad meal today, you probably won't tomorrow either (and if nothings changed then even a year from now it won't be more likely). I do this with food, outdoor activities, video games, you name it.

It's why I decided to pick up this book as this books primary focus is on efficiency, specifically the elimination and reduction of efficiency in exchange for effectiveness.


Imagine a secretary who is busy for only 4 hours out of the day. The rest of the time he sits there and maybe wanders the halls looking for something to do. Perhaps one day an engineer walks up as asks if he can enter this data into spreadsheet or make some copies. Then next week a manager asks to book individual meetings with her and all of her staff. Each of these items takes up maybe 2 hours of the secretaries time. This isn't very efficient.

A outside efficiency consultant comes along and notices that there are multiple secretaries for each group that only is busy 4 hours of the day on average. The efficiency consultant advises the company to combine all the secretaries duties so that every secretary is busy for 7.5 hours each day (careful to leave that extra 0.5 hour to handle the 2 hour tasks that seem to pop up once a week) and to fire all the rest. Now the secretary is busy all the time and when an engineer or manager has some work for him to do, they now find that they have to wait 4 days for the 2 hours of work to be finished. 

This is certainly more efficient, but not effective. You have gained efficiency but you have lost responsiveness. Imagine at your job you are 100% busy for a whole year. New technologies arrive or new techniques are invented, but you don't have time to discover or train or use these new advances which could help you do your job, as you are too busy. This is where slack comes in.

A penny saved is not always a penny earned, but it is always a penny not invested. Investing in slack (which can take many forms such as of time, control, or power) may not seem efficient, but it can be effective.

Lost But Making Good Time

You and your manager sit down and open up the latest project planning software as you begin to plan out and schedule the latest project you are responsible for. As you create the project it asks you a question.

Do you want to:
-Minimize time
-Minimize cost

"God damnit!" You manager barks out. " I want you to minimize cost and time!"

When a person is told to complete two mutually exclusive goals, the result is stress. Stress is a sign of slack deficiency. If the manager gave you slack on the amount of time you could spend, you could minimize the cost. However giving you no slack in either direction just means at best you'll accomplish some weighted compromise of the two, and at worst the extra stress will cause you to use bad judgement in some other area that in the long run costs you and the company even more.

Most slack averse companies are plagued by a culture of fear. When you realize the root cause of most frustrating corporate decisions isn't stupidity but fear, a whole lot of absurdities (like the one above) begin to make sense.

For example at my current job, I am plagued and drowned by standardized processes. People who are vastly removed from the work are required to sign off on documents or decisions they know nothing about. I have to go through 30 step processes which give me specific guidance on how to do all the mechanically easy stuff but give no guidance on doing the actual core or hard part of the problem. I used to pull at my hair trying to figure how these rules could be created. When you realize its fear driving decisions like this, it all of a sudden makes sense.

Process standardization is a direct result of fearful management. As a defense against failure, a standardized process is a kind of armor. The more worried you are, the more armor and standardization you pile on. But armor has a side effect of reducing your mobility. 

If you're facing your problem head on and its coming right for you, perhaps this isn't so bad. But what if your real threat is an unknown (as it is most of the time) and its to your left, or ten steps to your right. Well if you can't turn and face it in time for it to stab you in the side, then all that armor didn't do you any good.

Change and Growth

If you never allow your employees any slack time to fail, learn, train, and most importantly spend time that could yield no benefit, the best case scenario is your company stays exactly where it is. The more likely case is it regresses.

Who can drive change in an organization? Can it be pushed down from the top executives down onto the plebs? No. Can the front line workers drive change? No, they don't have the power. Change will be primarily driven by the middle managers. Be sure to cut them some slack.

Risk Management

"Best case scenario I will finish this project in a year, however given the risks which will cost a certain amount of time if each risk materializes, peak probability is at 18 months, with the tail falling to 0% probability of taking longer than 24 months."

"Well lets aim for 12 months! I can't wait to tell the higher ups!"

"Remember, 12 months is possible but not probable. Like I said, peak probability is at 18 months."

2 days later

"Hey boss, listen I found that if I have all my workers spend a month doing this specific training we can reduce the impact if any of the risks end up materializing (think practicing fire drills in school or placing fire extinguisher). If the risks never happen it will have been a waste but it shifts my peak probability to the left quite significantly. However it means that if everything goes well 13 months is now the minimum time to complete the project."

You can guess how that would go over with the boss who just loudly proclaimed 12 months to the higher ups. Attached is a picture showing the probability distributions. Without any slack in the schedule, you can see how a boss would make a decision against the company's interest.

Originally posted on Facebook April 2, 2017.