I rediscovered the site 99designs.com this past weekend. If you haven't taken a look, you should: it's a pitch-perfect application of crowdsourcing principles. And, as you browse the folios of some of the designers, you come to realize that there's a ton of design talent out there that's not attached to an agency or other enterprise.
It got me thinking about my employer (which--no surprise--derives a significant chunk of revenue from design services) and disruption curves. What prevents our clients (primarily F1000 companies) from crowdsourcing their interactive needs? What determines how quickly crowdsourcing approaches to a given knowledge worker job become "good enough" for enterprise use? How soon do my colleagues and I need to start thinking seriously about hanging their own virtual shingle?
While I haven't yet looked closely at the data or done the math, I get the sense that there are three (3) key variables that impact the rate of change to a crowdsourcing approach for a given job / business capability. Now I just need to find the effective data to prove my hypothesis:
Specialized Knowledge (positive correlation with rate of change): Take a look at the career disciplines that are trending more quickly towards croudsourcing: visual design, legal work, contract research, programming, writing / editing. They all require deep expertise, both to perform the work and to effectively evaluate work opportunities for fit. In addition, the knowledge worker "goods" they produce are created independently or with limited acquirer involvement, making crowdsouring models attractive for both the provider and the acquirer.
Knowledge work requiring lower levels of expertise (e.g., formal customer support, process fulfillment) has followed a separate path out of the enterprise, via outsourcing and offshoring. While there is definite movement towards decentralization of these jobs (e.g., work from home customer support), and solutions are beginning to emerge to enable task distribution and resource matching (e.g, Amazon's Mechanical Turk), the business infrastructure hasn't evolved to the point where these activities can be effectively crowdsourced at scale (my ballpark estimate: 2-4 years to inflection).
Personal Capital Investment (negative correlation with rate of change): I'm writing this post on a sub-$1000 laptop. Along with my mobile phone, I have all the tools I need to produce my knowledge "goods". From a capital investment perspective, mine is an easy job to crowdsource. A nuclear physicist, on the other hand, needs lasers and colliders and magnetic bottles to do their work. (Sidebar: why aren't more physicists evil supervillians? Money! It takes serious cash to build planet-obliterating particle beam satellites. But who wouldn't want to own one, even if your downfall at the hands of the debonair secret agent were assured? I digress...) Interestingly, when it comes to sciences, it's beginning to look like certain aspects of life science may have a shorter runway to crowdsourcing, given their nascent DIY movement (see yesterday's announcement about Petridish and today's NY Times piece regarding the rapidly decreasing costs for gene sequencing).
Perceived Value / Risk of crowdsourced "good" to the acquirer (negative correlation with rate of change): To date, the jobs / business capabilities exhibiting a faster rate of change to crowdsourced fulfillment are already being outsourced. Put another way, the aqcuiring organization already has an understanding of the value and risk associated with sourcing the "good" from its extended enterprise. 99designs.com is a good reference case: visual design services have been migrating out of the enterprise for decades: crowdsourcing is simply accelerating the movement by providing a newer, more efficient sourcing channel.
Of course, the counterargument to this claim serves up examples like P&G's "Connect + Develop" or the Open Innovation platform Innocentive, where the crowdsourced goods produced are integral to key business processes. While companies building virtual R&D capabilities are realizing significant value, the domain risk / value function is generally well understood: they've been leveraging contract research entities for years. At the same time, successful case studies like P&G lower perceived risk within the domain as market followers justify adoption as a means to maintain competitive position. Real risk diminishes in parallel, as the business and legal structures clearly defining responsibility and ownership are formalized.
My opinion is that there's something fundamentally human about the crowdsourcing ethos. Enough has already been said about the compromises and limitations of legacy work structures that evolved over the last century. Ironically, these structures serve neither the employees that live them nor the enterprises that enable them very well. Today's crowdsourcers are returning concepts of accomplishment and value realization to the individual, where they belong.
More updates as I dig into the dig into the data. Feedback welcome!
No comments:
Post a Comment