Savvy leaders are incorporating data science micro-jobbing (a variation of independent contracting) into their strategy. The key to pulling it off is to build the capability for micro-jobs.
The financial crisis of 2008 and 2009 not only jolted the world's faith in corporate stability, but it also forced people to explore different ways of generating income. This gave rise to the popularity of a variation of independent contracting called micro-jobbing and to platforms such as TaskRabbit and Elance that support this way of doing business. The difference between micro-jobbing and traditional contracting is the size of the job: Micro-jobs are very small in scope.
Micro-jobbing gives workers a way to earn money without making a long-term commitment to a company. It's a mutual benefit for employers, because there are situations when a company doesn't need or want a long-term commitment with a resource.
See benefits to your data science strategy
Micro-jobs can benefit your data science strategy in a number of ways. Micro-jobbing's most significant advantage over traditional contracting is that it avails you of a vast pool of talent you wouldn't otherwise be able to access. You might be surprised at the types of people who are learning data science these days; everyday people are aware of the demand in the marketplace for data scientists, and they're taking it upon themselves to learn information management and predictive analytics. That said, not everyone has the availability or the opportunity to jump right into a full-time data science job, so there's a lot of very bright people who can contribute to your organization given the right structure.
One benefit that might seem obvious though I don't recommend going heavy with this approach is: have a micro-jobber solve a tough problem that is stumping your data scientists. First of all, really tough problems are usually not small problems. More importantly, if someone can solve a problem your in-house data scientists cannot solve in a few weeks or so, you should reconsider your in-house data scientists.
A micro-jobber can, however, keep your existing data scientists on their toes. For instance, you might have a micro-jobber challenge a key theory to see if they can disprove it; or, you could have them redo a qualitative study on your market to see if they come up with different themes. These checks and balances keep everyone at their best.
Build the capability for micro-jobs
Although micro-jobbing can add huge benefits to your data science operation, it will require some organizational structure and development before you can pull it off properly. I encourage you to follow these three steps.
- Know how to decompose your scope. The scope on micro-jobs must be very small. Don't search for a micro-jobber to do a job that will realistically take three to six months of full-time work; this will not work for you or the micro-jobber. A typical job should last a few days to a few weeks.
- Know where to look for micro-jobbers. You could leverage one of the existing platforms, but I discourage this practice. Most of the commercial micro-jobbing platforms are targeted to attract low-fee work. (There is Kaggle, though it's not really a micro-jobbing site, and its competition model -- while relevant and effective -- isn't exactly what we're looking for here.) I suggest building your own micro-jobbing platform and keep it within your brand. It's not as difficult as you might think, given the available commercial platforms that you could model against. You'll have to market your micro-jobbing site to the data science community, but with all the targeted outlets these days, that's not very hard either.
- Work with your Procurement function to streamline your on-boarding process. When I negotiated my master services agreement with Visa, it took over two months from start to finish, and they requested $5 million in professional liability coverage -- this would never work in a micro-jobbing situation. You must have the systems in place to support very rapid on-boarding with reasonable requirements. Sometimes this is just a matter of having a discussion with Procurement on how they would handle a micro-jobbing scenario. In many cases, Procurement will take a one-size-fits-all approach to contractors. To be successful, Procurement can't treat a micro-jobber like a contractor who will be onsite for a year.
Micro-jobbing may be the way to take your organization to the next level. There's a huge, untapped resource pool of brilliant analysts that can contribute the intellect and creativity you need to achieve your vision -- you just need to know how to put the pieces together.
Ride the micro-jobbing trend by building your own micro-jobbing site, and then post small challenges out to moonlighting data scientists that need a little extra income -- this will bring fresh ideas into your organization and keep your in-house data scientists at the top of their game. Procurement is usually a sticking point, so run this scenario by them now and see what they say.
The micro-jobbing trend is already in motion to support your strategy -- you just need to hop on and take a ride.