This post is the eighteenth part of a ramble-rant about the software business. The current posts in this series are:
This post is about attempts to repeat once-in-a-project occurences.
...sometimes only once.
Between 1924 and 1932 experiments about human productivity were conducted at the Western Electric Hawthorne Works near Chicago. One set of experiments involved illumination and serves as the basis of the Hawthorne Effect. The experimenters tested the premise that more light would increase worker productivity. The results of the experiment supported this hypothesis - worker productivity increased measurably when more lighting was added.
The lighting was returned to normal after the experiment concluded while observers continued measuring productivity and a funny thing happened: worker productivity increased again.
The Cause of the Effect
There are books written on why this is so, and as many opinions as there are opinion-holders. I fall into the camp that attributes the effect to a combination of two things:
- The placebo effect.
Change, because anything done to increase attention communicates "We're interested." The placebo effect says the workers knew they were being measured, and the attention inspired or motivated them to do more or better work. Do you see the common thread?
Why is attention important? Attention communicates care and concern. It conveys importance. If the boss is worried about something or wants more (or more frequent) feedback, the line worker gets it - "this is hot and requires focus." The increase in productivity resulted from a change in demonstrated interest (increased attention), and not from the mechanism used to deliver the message (increased / decreased illumination).
Does this translate to knowledge work? Yes, but how do you sustain changing levels of interest?
Turn It Up
One way is to constantly increase the pressure on a project. This is easy if you have a deadline, but it is not sustainable for two reasons:
- The deadline will pass (hopefully with the desired work/project completed).
- Cranking up the heat (attention) yields asymptotic results (there is a finite amount of blood in the turnip).
Are there other ways to communicate attention without turning it up? Yep.
The Value of Change
You can alter the work environment in ways that don't punch developers in the brain. One way to accomplish this is to introduce pair-programming for some period of time. Another is to alter the location where work is being performed. Both changes will yield results the first time you try them. Diminishing returns will kick in, however, and the results of such tactics will prove as asymptotic as any other.
In other words, the boosts in productivity decrease with each re-use of a technique. This reduces the effectiveness of change as it's tried again and again over time. It's why I refer to these as "one-time boosts".
The Central Rule for Andy's Projects™
It's important to remember the Central Rule of Andy's Projects™: The energy pumped into a project is conserved, and people are the only medium in which this energy rests. This means the pressure applied to people working on a project remains with the project until the project completes. It may not be evident, but it's there until the end.
I remind readers who know (or work with) me that I attempt to place at least one year between what I write about and what's going on in my personal or business life. I add this reminder here because work has me traveling quite a bit lately as deadlines loom.