When it became obvious that we would not be able to complete LetMeGo on schedule, we chose to move the entire development team into a round-the-clock “LetMeGo Immersion”, in which the team lived together in a single house, and worked essentially every waking moment for 90 days.
This post is a brief summary of the costs and results of this effort.
The Challenge
LetMeGo is an ambitious project. In terms of software, it will take almost 27,000 man hours to reach the beta stage. Our original plans called for reaching the beta stage of the application in May of 2009, but by January it became obvious that we would not be able to stick to this schedule.
Time Being Productive (green) vs. Time Wasted in Communication (gray)
Most managers would delay the release or add more people in order to cope with this situation. We didn’t think that would work in our case. In projects with complex frameworks, doubling the size of the team will not double its productivity. In fact, I have found that in some cases the combined productivity of a team of seven is likely to be higher than that of a ten-person team.
We needed to find an alternative approach that would allow us to get back on track.
The Premise
In January of 2009 I met Eric Litman, a serial entrepreneur. I learned that, in order to release his newest venture on time, he hired seven developers to work 24×7 out of an apartment for 40 days. The premise made sense: Most of us expend the lion’s share of our time on sleeping, eating, commuting, and enjoying ourselves with our families and friends. Eric had created an environment where people could forget about most of those tasks and simply focus on work. If we could do the same, we could increase our productivity several times, keeping the team small but working more hours per day and per week.
The software behind LetMeGo is quite complex, and we were well behind schedule, so 40 days would have not been sufficient. We guessed that 90 days of immersive work would get us on track to release the beta version. We then worked to convince the current development team and to hire new people. It took a significant chunk of money and effort. This was quite a gamble for all of us. After all, while we would likely reap significant productivity gains, we could not be certain of all of the follow-on effects that such an environment might produce in the team.
The LetMeGo Immersion team: Maestro, Lucho, Andres, NIckman, Henry, German, and me, Alex
The Execution
I wrote many posts detailing the execution of the LetMeGo Immersion, including these:
- Intro to the LetMeGo Immersion
- Back in NYC for just over one day and for a very good reason…
- Just Listed: The LetMeGo Immersion House
The Cost
The total cost per month was three times the amount we customarily invested into development. 75% of the additional cost went to salaries. 25% of the additional cost paid for the logistics (the house, food, the cook, tech support, cleaning, etc).
It is more difficult to quantify the human cost, though we are pleased to note that there were no fatalities due to programmer-on-programmer violence. Some of us paid a toll, though: 
 Andres before the LetMeGo Immersion |
 Andres after the LetMeGo Immersion |
The Results
We didn’t reach the beta milestone completely, but we were able to release a limited beta version of the site. We are still developing some of LetMeGo’s traveler-focused features.
We found that the productivity of the team was also three times higher than usual. For those of you familiar with RescueTime, the tool we used, we measure productivity by multiplying average efficiency per week times the tracked time worked per week.
Productivity = Average Efficiency * Work Time
Productivity Working as Usual = 1.22 x 30.9h/week = 37.7/week
Productivity at the LetMeGo Immersion = 1.53 x 73.8h/week = 112.91/week
Ratio: 1:2.99
Given that the productivity and the cost increased side by side, we can then conclude that the biggest benefits is that we saved seven months of development. If it wasn’t for the Immersion, a limited beta of LetMeGo would have not been released yet and the full site would have not been launched until the second quarter of 2010.
Would I Do It Again?
Yesssss!
Alexander Torrenegra
P.S. I asked the team to write in one phrase what they think, in retrospect, of the LetMeGo Immersion. These are the answers I received:
The immersion was an strange experience mix, between ‘I must do it’ and ‘I will never do it again.’ I highly recommended it to single, work-addicted, geeky, and chatterbox men. German
The immersion was an experience that could have driven me crazy, but the great team ensured that I was just crazy enough to keep going and enjoy the most important challenge I have ever had. Lucho
A smile is drawn when you belong to an idea… Nicolas
The immersion was an excellent experience, with a high level of concentration, productivity, communication to solve issues, challenges, and desperation… I thank the Wii Fit for helping me alleviate tension and survive. Henry
Alexander is LetMeGo's co-founder and jefe. Here is his LetMeGo 
“Most managers would add more people in order to cope with this situation.”
Would they, now? I was rather hoping people would have heard of Fred Brooks by now… I say, 34 years should have been enough for the message to arrive!
Thank you Offh. You are right. I oversimplified my statement.
For other readers of the blog, Brooks law says that “adding manpower to a late software project makes it later”. Here is some additional information about it http://en.wikipedia.org/wiki/Brooks%27s_law
Yes, thanks for the clarification.
Anyway, you obviously avoided the trap of the mythical man-month… good for you! =)
I certainly hope that taking more efficient measures – such as you guys did – is not the exception any more, but the rule.
You certainly got cheap development that way – unless my math is wrong, this translates into roughly US$10/hour. Let’s just say I hope your developers got quite a bit of equity and you all make out like bandits
Strangely enough, it *still* fascinates me. The idea of being able to focus on a project 24/7 just has a certain draw to it…
Hi Rachel. It is actually more than US$10/hour. When you take into account the weakness of the dollar, currency exchange fees, the social security taxes in Colombia, vacation time, health insurance, etc., it adds up to almost US$20/hour (and that doesn’t take into account the stock options each team member is getting). That is still cheap compared to US standards, although it is a lot more than what you would pay in India or Pakistan.
Great job. I really admire the fact that you could eat sleep and breathe one project for the whole 90 days.
What where your “work hours” exactly? Did you have a schedule of some sort?
Bahaha, I just took a look at my post and the English seems horrid. The Redbull has taken its toll.
Cheers
Each person had its own schedule, but more or less we invested the same amount of time at each daily task. This was mine:
> 7:30 AM Wake up and start working right away
> 10 AM: 15 minute breakfast break
> 2 PM: 45 minute lunch break and then go back to work
> 5 PM: 30 minutes of exercise (Wii Fit)
> 5:30 PM: 30 minute shower and then go back to work
> 9 PM: 30 dinner break and then go back to work
> 11 PM: 20 minute break to videoconf with my lover, Tania, and then go back to work
> 1 to 2 AM: Go to bed
Rachel – he specified that development costs tripled. I assume this means they were paying their developers for all that overtime.
Ok, so I have to ask…
At the end of it, did you LetThemGo?
I’ve been following the progress of the LetMeGo immersion ever since my friend Ana Milena introduced me to your blog. Thank you for sharing the process and conclusions. Very interesting stuff! I think this is a great solution for huge projects on a tight schedule but I wonder how much can a person take this life style… can it be more than 90 days? How were the spirits of the group close to the end?
90 days sounds like a lot for us, developers, but there are many other professions where 90 days immersed at a job isn’t that much. For example: oil rig engineers. I think that another immersion could last more than 90 days, but you must make sure the team members can handle it. You can do that by only hiring people that have had similar experiences before.
The spirit of the team was quite similar from beginning to end. At the end, of course, we all wanted time to pass quickly, but that didn’t affect the productivity or the atmosphere significantly.
From the job offer: “The stock option bonus will be effective if LetMeGo.com reaches beta before June 6th.”
From above: “We didn’t reach the beta milestone completely, but we were able to release a limited beta version of the site.”
So did they get the stock options? Or were the programmers paid just the USD$14,520 for the three months of work?
They delay wasn’t their fault. As such, they got the stock options.
Was there time/productivity lost after the immersion push?
Normally, a break is needed after that sort of thing.
People who work on ships (in well off countries) work 3 or 4 months on and then have two or three months off (paid).
Otherwise, long term productivity goes down.
Hi Andrew. A few of us took a short five day vacation to recover. Other have long vacations scheduled right after the full launch of LetMeGo. The productivity was low the first few weeks after the immersion ended, but we are now back on track.
Por curiosidad, en que parte de Bogota hicieron esto?
Here is the map: http://letmego.to/immersion
Very nice!
This is great! It’s too bad we can’t do this all the time.
I’d like to know more about your software development process. Did you do any user research? Was your design user or activity centered? Any designers in the team?
Hi Bernardo. It is all described here: http://torrenegralabs.com/group/content/rrapido_methodology
Yes, there are designers in the team. We do have done and continue doing user testing to tweak the interfaces we have developed. The design is usually the result of having determine what should be shown to the user when, depending on the use case. I hope you find this information useful.