Welcome to the IdeaMatt blog!

My rebooted blog on tech, creative ideas, digital citizenship, and life as an experiment.


How I wasted two years on Twitter, all because I wasn't tracking

[Cross-posted to Quantified Self]

Guatemala sinkhole 2010

Between 2007 and 2009 I spent a ton of time in Twitter before it finally hit me that 1) the net improvement to my life was zilch, and 2) had I thought of it going in as an experiment, I would have quit a long time ago and freed up energy for more effective efforts. Of course social media tools can provide plenty of value, but, as Alex said, Social media is an addictive time suck.

How do we go about measuring the value of Twitter? Business calls it ROI, but I think of it as simply what you hope to get out of it. The key is deciding why you're using it. In my case I was dabbling, which is a fine motivation, as long as it's done experimentally. After all, how many discoveries came from just getting curious and trying out something new? But here I should have set a time limit, and I'd still want to have something quantified, even it it's as soft as "perceived value."

But for more specific uses, coming up with measures is important. Are you trying to get more customers? Do you want to hear from people who can give you ideas for your product or book? Or maybe it's more of a social pulse use - keeping in touch. Some metrics are straightforward, such as # inquiries about your business, or number of tweets from others that made you smile. However, I think a major challenge is latency - the time delay between action on your part and resulting effects seen in your life. For example, it might be months before you hear from someone who's been silently reading your tweets. Maybe in those cases we could make the measure more direct by asking them explicitly what the impact is. I'm not sure.

While I didn't treat using Twitter as an experiment per se, I managed a few times to use Twitter itself as a platform for experimentation. For example, I had a business trip to North Carolina coming up, so I tested using the tool to make business contacts for the trip. I subscribed to a bunch of local folks, followed their tweets, and replied when appropriate. The result (measured in # business meetings set up) was zero, but the "at least now I know" feeling was satisfying. One thing that's vexing is irreversibility. Any experiments we might do with our followers has permanent effects because individuals have memory ("Oh yea, isn't he the guy who tried giving away the iPads?"), and the account has memory (I might loose or gain followers s a result). One workaround is to make separate accounts for testing, but unfortunately you can't clone followers. Then again, making us start from scratch is probably a good way to keep the experiment clean.

Others have used Twitter this way, such as a much-publicized psychic experiment. (I found the comments on Richard Wiseman's Blog announcing the experiment to be fascinating from the experimental design perspective.) A little searching turned up other posts like Twitter for Research, The start-up chronicles: Experiments with Twitter, Using Twitter for Market Research, and a Twitter Tip Sheet for Experiments. I don't think I'd count The Anybody/Everybody Twitter Experiment as one, though; what was measured?

What do you think? Have you used Twitter in one of these ways, i.e., as itself an experiment, or as a platform for running them? And if you regularly tweet, I'm curious to hear what you get out of it, and what you've measured to determine that.


"I can live with doubt, and uncertainty, and not knowing..."

I can live with doubt, and uncertainty, and not knowing. I think it's much more interesting to live not knowing than to have answers which might be wrong. I have approximate answers, and possible beliefs, and different degrees of certainty about different things, but I’m not absolutely sure of anything, and in many things I don’t know anything about, such as whether it means anything to ask why we’re here, and what the question might mean. I might think about a little, but if I can’t figure it out, then I go to something else. But I don’t have to know an answer. I don’t feel frightened by not knowing things, by being lost in a mysterious universe without having any purpose, which is the way it really is, as far as I can tell, possibly. It doesn’t frighten me.


Richard Feynman from The Pleasure of Finding Things Out: The Best Short Works of Richard P. Feynman


How contagious is courtesy? An informal experiment

Steward at Council even hold doors open for Commissioners

Like most clichés, "courtesy is contagious" had no real meaning for me until I experienced it. In this case, as I was driving I let someone enter a line in front of me, and noticed that next person behind me did the same. This made me wonder how contagious courtesy really is, so I decided to experiment with it a little.

First, I tried the opposite driving behavior - the more common zero-sum "someone else can let you in" approach where I did not let people into line. After trying this a handful of times and comparing it to the courteous case, I noticed that roughly twice as many cars passed before letting the waiting person in (the best I could judge from my rear-view mirror.) In other words, if I let someone in, then the person behind me one or two back would let in the next person in line. If I was not courteous in this way, however, it looked like ~3-4 cars back let them in. This is a poorly-controlled test, but not only did it seem to help, it made me feel better. Result: A keeper!

I've noticed a similar "follow the leader" behavior from the bicyclist perspective. Where I live in New England, we have narrow roads in the country, which makes biking both beautiful and dangerous [1]. What I observed is that when a driver gives my wide berth (actually crossing the dividing line - thank you!) then following drivers do the same. The opposite is generally true too: If someone whizzes by one foot away at 45 MPH, so do the drivers behind him. I still haven't figured out experiments to try from the bike to change this behavior. Angry waving doesn't communicate.

Another little example I noticed is while waiting in line at the library, I heard the elderly man in line before me tell the check-out person, "I appreciate all your hard work in the library." Being reminded of civilized behavior encouraged me to do the same, like scooching my chair back against the table I was working at for the next person.

Here's an experiment idea: For a week, actively look for examples of (un)common courtesy. Or better yet, pick a small courteous activity you can try out, do it fora a week, and stick around to see if anyone repeats. Not conclusive, but a bit of evidence, and plus it feels good. Who knows, you might start a trend!

I'm curious

  • Have you seen something like this in action?
  • Are you naturally courteous?
  • If you experiment with this, what were your results?


  • [1] A two-by-two, anyone? Beauty vs. danger (help me here):
    0,0: Ugly duckling
    0,1: The Stone Fish
    1,0: Disney princess
    1,1: Femme fatale

Just do it? But HOW? 24 productivity experiments I tried, plus a QS time management recap

In the workshop

[Cross-posted to Quantified Self]

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:


2011-03-11: They did WHAT?

Test Tube Terrarium

Quick links from the past week of experiments in the World Wide Lab

The Fish Pond: Vazquez's velocity increasing: I don't follow sports, but "experiment" comes up constantly in that context. In this case, pitcher Javier Vázquez is reported to have changed his form to use his lower body more. It seems to be working: "On Wednesday against the Nationals, Vazquez was clocked at 90 and 91 mph, after he was about 87 mph in his first outing." What's interesting is that experimentation is prevalent in sports (and probably has been forever), and that there are straightforward metrics for evaluating impact.

Facebook experiments with streaming movies: In "Tell me, Betty. Has your husband always been..." I mentioned Amazon's recent addition of instant video to their Prime program. Well apparently Facebook and Warner Bros. Pictures are getting into the spirit with an experiment to offer The Dark Knight for 30 Facebook credits (AKA $3). From the Think, Try, Learn perspective, this highlights how experiments can impact others. In this case, Netflix Investors Spooked By Facebook Video Experiment. Of course our personal and group experiments don't have this kind of massive impact, but they do affect those around us. Relationship experiments in particular come to mind.

Plastic-free experiment an eye-opener for Kanata family: Here's an experiment I admire, and that anyone could try. A family of four stopped buying things containing plastic. Wow! Given that nearly everything has - or is wrapped in - plastic, it's a challenge. The article mentions what we eat, what we wear, and how we wash and brush our teeth. It also shows the power of experimentation to change behavior, in this case naturally using much less plastic.

I was pleased to find recent stories on killing your TV (or at least not paying for it), such as What it's like to cut the cord and stop paying for TV and Savings Experiment: Cutting the Cord on Cable's Pricey Monthly Bill: We've not had cable for years, which I admit gives me an embarrassing and smug sense of superiority, along with the more humble satisfaction of not rewarding crappy programming by paying for it. That said, for us the experiment wasn't difficult because we haven't spent the time to find a series that we like.

In this Book review of The Science of Kissing: What Our Lips Are Telling Us, the author mentions some fascinating experiments, such as the 1995 "sweaty T-shirt experiment":

By getting 44 men to wear T-shirts without deodorant for two nights and then having women sniff them, Wedekind discovered that "women nearly always preferred the scents of T-shirts worn by men with MHC genes different from their own - suggesting that we can determine our genetic compatibility with potential partners simply by following our noses."

What I particularly like is the comment that the best science experiment is the kind that defies your expectations. Edison experiment, anyone?

Finally, what list of experiments could fail to mention Amy Chua's provocative Battle Hymn of the Tiger Mother (covered in Time's Tiger Mom: Amy Chua Parenting Memoir Raises American Fears)? To me, of course, its potential for experiments leaps out. So, I was pleased to see someone trying it out: Valerie Frankel: My Tiger Mother Experiment: Using Chua's Book as a Parenting Guide. Do you have any thoughts on this one? I haven't studied it yet.