I think about business a lot, either analyzing and breaking down a business that is serving me or thinking about my own business. I’ve done that for quite sometime. Around 1999 or 2001, I hatched an idea: Someday huge corporations will be replaced by revolving teams of contractors. Employees will be expensive and pointless. Instead, you’ll just announce a need and pay those that deliver. Workers will no longer want or care to work for anyone but themselves. They’ll prefer to work on projects they pick with people they like vs risking their future on one single corporation.
My idea was crowdsourcing before the term was coined in 2006.
It occurred to me that business as it stood back then was flawed. It started out good, i.e. you wanted to farm out the work to a set of experts vs trying to futilely build it in house. Very quickly though the process failed. You then had to seek proposals, review each proposal, each proposal’s owner(s) had to be vetted, then a time frame was worked out, budgets setup, etc. All of which meant nothing because every project I ever saw in corporate America would go over budget and scope.
My idea was to skip the proposal and vetting stage. Let ‘s face it. Most consultancy firms hire independent contractors anyways, i.e. you end up paying a markup of 50% to 100% just so Company A or Company B can hire Contractor A. Why not just get to Contractor A and his team for the project? Skip the middle man, go straight to the source. That’s what business is all about!
If you have a budget of $250K for a project, announce that project and the $250K cash prize for the team that delivers. This is where people don’t like crowdsourcing. Let’s say 10 teams sign up and start building. Only 1 team is going to win, so the other 9 will have wasted their time. “Who will pay for their time and energy? It’s unfair that they worked for free.” But you know what…who cares. Nobody cried for Company B in the past when Company A kept winning all the bids to hire the same contractors. To be honest, the goal of the paying company is to care for itself and it’s project. They need to be successful and on budget. This will in turn help it better serve it’s customers.
A lot of people have told me. “It’ll never fly.” Some will grant me some semblance of possibility, “Well, maybe for logos and small stuff like that. It’ll never work for anything big though.” By far, the most common negative response is, “Yeah, maybe for certain projects. However, you’d never do it for things that are important to your business. Those will always be done traditionally and inhouse.”
To all you nay-sayers and haterz out there, I present this little nugget for you. Netflix agrees with me and not with you. Granted Netflix still has regular employees, so it’s not 100% my vision. (Give it time, business moves slow.) However, it entrusted the improvement of its recommendation engine/algorithmn to crowdsourcing. They said, “$1M USD to the first team that can improve our recommendations by more that 10%.”
I can hear the voice of dissent among you. “Yeah, well they’re rich and can afford it. But what about the others teams? Also, so does Netflix now own their work?” Here’s the amazing thing about this prize (and proves that Netflix really gets it). This blurb is from the official rules page:
- To win and take home [the] prize…, you must share your method with (and non-exclusively license it to) Netflix, and you must describe to the world how you did it and why it works.
Hmm…did they misprint that? Not only does Netflix not exclusively own the rights, but now the winners have to share the knowledge with the entire world? Wait, this would mean the teams competing against them and Netflix’s competitors? Yup. I haven’t read Netflix’s reasoning behind it, but since I think like them, I can guess as to why:
- To shut up the anti-crowdsource whiners – Now, when Netflix decides to do this again. All teams will start on equal footing. All those participants that truly care about this crazy algorithm benefit from the competition exercise. Sure they may not have gotten paid money, but they gained something far more valuable: knowledge. Knowledge of working together, finding out if they’re a good fit, etc.
- To still better the engine more – If the algorithm gets released into the wild, others can build upon it. The more people build upon it, the more benefits Netflix gets from it. Like any research, the value is from the peer review and building on top of, not the actual base research itself.
That’s it for me. I rarely get to say, “Told ya so” so I wanted to take the time to do it for this one. :) Bravo, Netflix and BellKor’s Pragmatic Chaos (the winning team). I hope to see more work done like this in the future. It makes me proud to be in business.