Web Ben's Questions and Answers

Friday, April 29, 2005

Software Crab Fishing Style

I saw a couple of episodes of a reality TV show about the people who are involved in the Snow Crab fishing industry in Alaska, in particular in the Bering sea. The programme was very interesting and I have to admit that I was certainly feeling tempted to try it and had considered the idea briefly back in College. As a result I began to ask myself; "What is so appealing about Alaska Snow Crab fishing?".

After reflecting on that question for a couple of days I've come to the conclusion that it's partly the excitement of the potentially high stakes, it's certainly a dangerous job, but mostly I like the idea of being rewarded based on your performance. The Alaska Snow Crab fishermen, from deckhand to captain, all get a share of the catch. The "Green Horn" first timers are the only exception as they are paid a fixed rate as they are so much less experienced they're really only getting cost of living compensation for their apprenticeship. The Alaska Snow Crab season is variable length, depending upon the estimated stocks of Crab, and now only lasts a few days. Due to the time constraints the Crab fishing season is a virtual sprint between crews who each want to catch as large a portion of the quota as possible before the season end date is announced. Consequently crews work 24 hours a day for these few days each year, catching sleep in 15 minute intervals whenever they can. However, the crews are paid well for this adrenaline rush and their take home pay at the end of the brief season may be as much as $300 per hour. Considering that the crew is working 24 hours a day during this time you can easilly see that financially the reward is very significant.

Since my business involved computers, I was wondering if the same model which is used in the Snow Crab fishing industry would work for the Computer Consulting/Services Industry. Several people who I have spoken to about this believe that the Computer Consulting Industry is already operating under the same model, however, I'm not so sure. Certainly, contractors are paid hourly, however, that's actually the opposite of the Crab fishing model. The Crab fishermen are paid a percentage of the boat's catch revenue, if they want more money they need to work harder and faster, since the revenue is limited by the short duration of the season. To mirror this in the Computing Industry it would be like a firm offering a bounty for the delivery of a computer system before a certain set date. Multiple independent teams of contractors would each compete for the bounty by delivering their version of the computer system specified in the firm's bounty requirements document. At the time of delivery each solution would be evaluated and the firm would choose how to distribute the bounty, a single team may receive the entire purse or it may be distributed amongst multiple worthy teams. Admittedly this scenario would be slightly different than the case with Crab fishing.

The real appeal of this model is the fact that it is largely self regulating. If a firm offered a bounty but no teams came forward to meet the challenge then the firm would be free to adjust the value of the bounty upwards or the date of delivery outward until teams did sign up. A team would be free to either drop out or join the race at any time and a firm would be free to adjust the bounty and the delivery date at any time. The most successful team would tend to be the one which worked hardest for the longest time, since a team would want to deliver the best solution possible at the time that the firm judges the solutions. If a team was confident in their solution they would be free to deliver it early, and a firm would be free to choose to reassign the delivery date and settle for the early solution if they choose. The consulting team would then be able to pursue other projects, however, they run the risk that their solution may not be chosen if the firm decides to wait until the original delivery date for competing solutions before choosing how to distribute the bounty. If a firm was unhappy with the solutions submitted at the time of delivery they would be free to set a new bounty and delivery date for solutions to remedy the failings of the current submissions.

In order for this model to work there would need to be an open ranking system for both individual team members and firms. When a firm received an application to compete for a bounty from a consulting team the application would include the teams past performance/satisfaction grade. This grade would be calculated as a sum of each team member's performance grade, which would be calculated from the team member's past peer review based grades combined with the team member's past team's firm ranked solution grades. This way a firm would be able to decide if their bounty and delivery date were likely to receive solutions from teams which had previously provided satisfactory results. If only mediocre teams apply to compete for the bounty then the firm may choose to adjust the criteria for the solution either though changing the requirements, changing the bounty, or changing the delivery date. When a team receives their share of the bounty they would grade the firm, which would enable a team to choose which bounties to apply to compete for based not only on the bounty solution criteria but also on the firm's standing in the eyes of those who have been involved in projects for that firm in the past. In other words; firms would grade solutions, team members would grade firms, and team members would grade each other.

The benefit of this system for the firms is that they receive the best possible solution based upon their requirements, their bounty, and their delivery date.

The benefit of this system for the individual team members is that they would receive a portion of the bounty according to their their skills and effort as evaluated by their peers, i.e. their team mates, and their customers, i.e. the firms when they evaluate the team's solution.

What do you think? Please add a comment or send a question to Ben Dash at


Post a Comment

<< Home