Matthew Cornell

March 18, 2011

Some time ago I was asked for the ultimate productivity tip, and instead of giving a straightforward take-away, I said that in the end the answer is “it depends.” That wasn’t a cheap shot because what works for you might not work for the next guy, and vice versa. Sound familiar? It’s the same case for medications, meditation, and most anything else we humans do. That’s why it’s best to experiment, examine your results, and decide based on the data. In other words, quantify!

But there’s a complication. Coming up with metrics that reflect the value of what we do, rather than the individual efforts, can be a challenge. While the latter are simpler to measure, (there’s a reason that some jobs require you to clock in – “seat time” is an easy metric), the real test is more how effective we are, not just how efficient. I may be cranking widgets at a fast pace, but what if I’m making the wrong ones?

Until we have general-purpose and quantified framework for measuring value (“accomplishment units?”), we have to keep being creative. In this long post I want to seed some discussion by sharing two things: some specific productivity experiments I’ve tried, with their results, and a recap of the cool productivity experiments found here on Quantified Self. Please share techniques that you’ve found helpful.

Productivity experiments I’ve tried

Adopt a system. The single biggest productivity change I made was trying a system for organizing my work. In my case I got the GTD fever (Getting Things Done), and my results were clear, including getting far more done more efficiently, feeling more in control, and freeing up brainpower for the big picture. At the time (five years ago) I wasn’t thinking of it in terms of an experiment, but it certainly qualified. From a QS perspective it can function as a kind of tracking platform because it has you keep a comprehensive and current list of tasks (Allen calls them “actions”). I have used them for various tracking activities, mainly by characterizing or counting them.

Two-by-two charting. I’ve plotted 2D graphs of various task dimensions to analyze my state of affairs, such as importance vs. fun (a sample is here). These are a kind of concrete snapshot that I analyze over time. In the above example I decided that the upper right quadrant (vital + fun) was still a little sparse.

Artificial deadline. A standard productivity idea is to impose a deadline on “as soon as possible” tasks that don’t have a hard completion date. For two weeks I tried assigning one to each task, and prioritizing based on that. It was a disaster, and I hated it. I think it’s a function of my personality, but my stress level shot through the roof.

Daily planning. The idea of creating a fresh “to do” list for each day is a classic time management idea, so I tried it for a month. Each day I’d pull out a mix of tasks (e.g., important, fun, large, small) that I thought I could get finished, then work strictly from that list along with my calendar. The test was how many tasks I got done, along with overall feeling of accomplishment, as compared with working directly from the master list (70+ items for me at the time). The result was huge. I got a larger percentage of tasks done each day (about 25% more), mainly because I had a kind of “focus anchor” that kept me moving. Also, it made it easy to turn it into a game where I tried to get a “touchdown” – all of them checked off. This changed my behavior to be more realistic about my expectations, which was both painful and satisfying. You can find an example daily plan here.

Estimated vs. actual completion times. For a few weeks I tried estimating how much time tasks would take, timing them, and comparing the two. Not surprisingly, my estimates were overly optimistic in general. (Apparently most of us have trouble – see the Wikipedia article on the Planning fallacy, esp. the paper “Exploring the planning fallacy: why people underestimate their task completion times”.) The above daily plan example shows using the daily plan to track it.

Tiny tasks. Another thing I experimented with was breaking tasks into a maximum of 10 minute chunks. This is an extreme example of divide and conquer, but it helps avoid procrastination because a small task wasn’t as much a turnoff as a giant one, making it easier to get started. Not only was this actually the case, I also found that once I got into the flow I was usually able to get more than 10 minutes of work done.

Interruption log. Because interruptions are a common productivity drain, I thought I’d try an experiment of simply tracking them. I recorded who interrupted me, when it happened, the topic, and how long it lasted. I did this for a week and then studied it for patterns that I could use for prevention. By far the most common type was internal: my frequent multasking, esp. checking email (more below). I found the simple act of tracking was an embarrassing eye-opener, and I’ve since been able to manage it via further specific experiments.

Task input/output. For a few weeks I measured how my tasks list size changed over time, looking for I/O trends. By recording starting and ending counts each day, I learned that I had a small “leak” where I was accepting more commitments than I could get done. This helped me get choosier about what I said yes to, giving me a good personal excuse to turn down work. Putting items on a “not doing” list helped relieve the guilt.

Incremental vs. batch. At one point, tracking expenses became a unpleasant time sink, so I experimented with doing incremental processing of receipts once a week, rather than waiting for the monthly bank statement to do them. The result was a massive improvement. The latter used to take about an hour, but the incremental combination was about 15 minutes (5 minutes per week entering, and 10 minutes reconciling).

Email. Email deserves a special experimentation category because it’s such a black hole for me. If I’m not careful I spend too much of time “checking,” rather than actually working through messages. (I work from a zero base, i.e., don’t stop processing until the inbox is empty.) Here are some experiments I tried, with a quick thumbs up/down:

• No Thanks rule (i.e., don’t send an email whose only purpose is saying thank you). Thumbs up.
• Use a Mac dashboard widget for monitoring gmail every five minutes (gasp!) Thumbs down.
• Create a Firefox smart keyword to search Gmail without having to look at the inbox. Thumbs up.
• Adopt a 24 hour maximum email response time (allow up to a day to respond). Huge thumbs up!
• Turn off the new mail annunciator. Another big thumbs up.

I’ll finish with some other productivity experiments I’ve tried:

• Used the STING method to stop procrastinating (Select one task, Time yourself, Ignore everything else, No breaks, Give yourself a reward).
• Do one High Value Task a day.
• Process a book in one hour.
• Used a digital voice recorder to take reading notes.
• Outsource transcription of dictated notes.
• Schedule tasks on calendar (no lists).
• Use time blocking (schedule regular chunks of time with myself for important tasks).
• Work in one big window (i.e., blank computer background).
• Disable certain sites during project work (especially Gmail).
• Change work venue (conference room, and cafe).

Recap of productivity experiments shared here on Quantified Self

There’s a lot of experience, advice, and ideas about productivity here on QS. Following is a sampling of the creative techniques that people have tried out.

Gary’s comment on my post Discuss: The Quantified Worker listed these (his words):

• Word count: this is simple but powerful; Buster Benson is supplying a nice general tool for this at 750 words. I track them in my word processor.
• Time arrived at my desk: I have done this for long stretches at a time, when subject to competing demands. Simple pattern awareness very helpful.
• Total work hours: This is not as helpful to me, as quality varies, and increasing “time at desk” does not always increase quality/quantity of work.
• Time in total concentration, 25 minute intervals: I picked this up from Robin Barooah’s talk on his caffeine and productivity data. (See The false god of coffee.)
• Time in concentration, 6 minute intervals. In this case, you keep a timer running and make marks on a piece of paper for every six minutes that passes. The first four marks are dots that form the corners of a square, and the next six marks are lines that connect the dots, completing the outline of the square with diagonals crossing in the middle. Total: ten marks = 1 hour concentrated work.

Kevin’s post “Productivity” Dashboard Monitor talks mentions RescueTime as being a kind of productivity information center that helps to analyze your tasks. You can find quite a list of similar tools in the Productivity section of Self-tracking links to get you started, such as Harvest, TimeDoctor, and Slife.

Gary’s writeup of Ping’s Thesis – From Diary to Graph shows a lovely graph of data from a widget Ka-Ping Yee wrote to log his activity.

Finally, you might want to watch these related QS videos:

• In Colin Schiller on Time Management Experiments, Colin talks about experimenting with using the Pomodoro Technique and only working eight hours a day. For four weeks, he tracked all the work activities he did in each 25-minute work segment.
• The video Peter Robinett on Predicting Productivity describes his simple spreadsheet-based system to track, visualize, and predict productivity, including color coding activities.
• At Bill Jarrold on Productivity Tracking, Bill shares his analysis of UNIX command line logs (a novel, if specialized idea) along with his results.

[Image from By Elsie esq.]

(Matt is a terminally-curious ex-NASA engineer and avid self-experimenter. His projects include developing the Think, Try, Learn philosophy, creating the Edison experimenter’s journal, and writing at his blog, The Experiment-Driven Life. Give him a holler at matt@matthewcornell.org)