Bee on Extreme Productivity (QS 2015)

Tuesday, September 8, 2015
By bsoule

Slide from Bee's talk: Code. Eat. Sleep.

Once again our esteemed cofounder, Bethany Soule, aka the Bee in Beeminder, gave a talk at the Global Quantified Self conference in San Francisco. But you don’t have to take our word for it. We now have video proof. We’re also sharing the transcript and slides, in case anyone cares that much. It’s based on Bee’s original blog post about her first maniac week but has some new insights (not to mention video!) so we’re blogging it again. See also our previous Quantified Self talks (and an interview): 2011, 2012, and 2013. For a real blast of nostalgia, check out Bee talking about Beeminder (then our pet side project, Kibotzer) at a New York Quantified Self meetup in 2009. And there’s also Bee and Danny at another New York Quantified Self meetup in 2010 talking about dogfooding an early version of Beeminder (still a year before we publicly launched, somehow) with $2000 contracts that we auctioned off to our friends.

Hi! Thanks for coming to my talk on extreme productivity! I’m cofounder and CTO of Beeminder, which is itself a kind of extreme productivity tool. I use it to drive a lot of my own productivity methods. Beeminder is a quantified self tool that uses your data to fuel behavior change. You pledge money to stay on track towards your goals, we collect your data, send you reminders, make a graph, and if you don’t do what you said you were going to do, we take your money.

We needed a tool to combat procrastination, and came up with Beeminder. There’s been a lot of experimentation along the way.

A couple years ago my cofounder, Danny, showed me Nick Winter’s blog post “The 120-Hour Workweek”.

Nick did 120 hours of coding in one maniac week by, as close as possible, doing literally nothing but coding and sleeping. That involved 6.38 hours of sleep per day, seven bars of dark chocolate, 402 unanswered emails, 44 commits, and 99 Trello cards slain.

I read about this and naturally I thought “That’s insane. … I want to try it!” Last fall I finally gave it a try.

For a “proper” maniac week you need to do more than just try hard to keep focused.

1. Publicly precommit

This lets everyone know you’re essentially off the grid, and gives you a strong incentive to stay focused. I announced it in a weekly Beeminder newsletter.

2. Remove distractions

Take vacation from your job, send your kids to summer camp, get an Airbnb in the middle of nowhere. Whatever it takes. I sent my kids to Canada for a family reunion.

Finally, the most important thing:

3. Document it

Take screenshots of your work, and of your face hopefully studiously looking at that work. Create a timelapse video and post it to the public record for awesomeness and glory afterwards.

Bonus points for writing a blog post.


I really enjoyed this, so much so that I followed it up with three more variations in the next 6 months.

But let’s see some stats: I more than doubled my usual 40 hours per week, putting in 87 hours. That’s 87 hours unambiguously on task, not just hours at the computer. I pushed 69 commits to GitHub, compared to 16 in the previous week, closed 19 Trello cards, and did 37 hours worth of pomodoros.

I got more than double the work of an average week done, and I enjoyed it.

What about efficiency? Programmers talk about task switching being costly, as you have to load a new context into your head every time you switch or are interrupted. This feels true to me, though I’m not sure what metric to measure it by. But big uninterrupted blocks of time feel valuable.

Most negative reactions go roughly, “Idiot. You’ll burn out!”

I’d burn out doing this every week, but I’m not! And in fact there were things that were really relaxing about it, like getting to sleep as much as I wanted every night.

There were hard parts. It was physically hard to sit still that much! A standing workstation would’ve been clutch. I also was surprised to find it mentally taxing to be alone for an entire week. A couple of my later iterations I had company working alongside me and those I found more pleasant.

The biggest failure or hardest part was insufficient preparation. I didn’t have a well developed task list. Deciding what to work on next was costly. Or a task I wanted to work on didn’t have a solid spec to work from. I’ve not solved the spec problem, but I’ve been using Mark Forster’s Final Version system to prioritize my task list lately. Recommended.


What else did I learn? What can be captured and brought back to my day-to-day?

Commitment matters

For obvious reasons, the screenshot every minute, and the public commitment to publish it after the fact were a huge pressure to keep working.

Normal people do this by, say, having a boss, or pair programming, or agile-style stand-up meetings. I use Beeminder. For example, our User-Visible Improvements goal promises $1000 to a user if we fall behind.

Don’t read email

It was amazingly attention-freeing to sign out of Gmail and know that no responses were expected for an entire week. The internet has a lot to say about strategies for dealing with email. (I’ve tried The Email Game and our own GmailZero.)

Looking at my data, it turns out I only spend about an hour per day in Gmail on average, so rather than being more efficient I decided to just care less.

Live at your office

I kid. But time spent commuting is irretrievable and having no friction to beginning my workday gave the whole day a sense of momentum.

Short of actually moving into your office, you could start the day by doing a pomodoro before breakfast.

Remove distractions

A Maniac Week is an extreme way to get rid of distractions, by effectively going off the grid for a week.

But in the day-to-day I’m my own biggest distraction — email, facebook, I think my phone just flashed, am I hungry?, etc.

Turns out there’s a lot you can do for this kind of distraction. For starters you can forcefully block yourself from distracting sites on the internet. RescueTime, StayFocusd, and Freedom are a few programs that do this. Or the old /etc/hosts trick.

Usually when I get distracted by the internet it is a lack of attention more than anything else. I finish a task and on autopilot I launch Facebook and I’m down the rabbit hole. Noticing this, the tiniest little fix has worked: I just sign out of Facebook when I close the site. Next time I launch Facebook, I get a login screen instead of the news feed. 98% of the time that’s enough to stop me.

Another tiny interrupt like this that makes a big difference for me is to move my phone out of sight and out of reach. It’s suddenly much less distracting.


The crux of the Pomodoro System is to set a timer for a manageable amount of time (canonically 25 minutes, but I like 45), and do focused work until the timer goes off. Then you take a brief break, rinse and repeat.

My most common failure mode with pomodoros is failing to start the next one. So of course I beemind it, and that keeps my momentum going. During my first maniac week, as an added incentive to my public precommitment I set a steep daily pomodoro goal for myself, and found that it was really effective at keeping me rolling through the day. Since then I’ve continued to track my total time working, but it’s the pomodoro goal that drives my productivity.

If “do focused work” sounds like the hard part, here are a couple tips:

  1. Pick a concrete task. (E.g., instead of “read”, use “finish reading 20 pages”.)
  2. Even better, write it down, or announce it publicly.
  3. If you notice you’re distracted, note down the distraction (literally, on a piece of paper) and then gently guide your attention back to the task at hand. (It’s like meditation, and you get better at it with practice.)

If you want to talk about Pomodoro Poker I have office hours and a break-out session tomorrow. Come on by to talk about extreme pomodoros. [Or, reading this after the fact, check out our explanation of Pomodoro Poker here on the blog. Or even join us for it if you’re in Portland.]

Tags: , , , , , , ,

  • robfelty

    Great talk Bethany! Thanks for sharing.

    You mentioned this briefly, but I will second the notion of doing group Pomodoros. I’ve done this with you and Danny in the past, and it is a great way to get work done, and do a bit of socializing too. Wish I lived closer so I could come to your hack nights.

  • Ron A.

    I like the notion of a Maniac Week, but I wonder how one’s productivity (in the sense of accomplishments divided by hours) would fare when compared to a more conventional schedule. I used to have a job where I worked late into the night, and there was a noticeable drop-off in everybody’s mood and work done after 10-11 hours.

  • Justin

    Curious about auctioning off your contracts to friends. Did any of the bid prices surprise you?

  • Daniel Reeves

    As implied by the no-trade theorem (or Aumann’s agreement theorem or the efficient markets hypothesis), it was hard to get people to bid much, even for a $2k contract. They figured (wrongly, I would say, given the close calls I recall we had) that we wouldn’t risk $2k without being damn sure the risk was minuscule. I think we sold the contracts for around $50. Here’s the hilarious interface we made for auctioning off Melanie’s $600 weight loss contract in 2010, which went for $12.32: