The promise – and perils – of hyperspecialisation

The promise – and perils – of hyperspecialisation

Much of the prosperity our world now enjoys comes from the productivity gains of dividing work into ever smaller tasks performed by ever more specialized workers. Today, thanks to the rise of knowledge work and communications technology, this subdivision of labour has advanced to a point where the next difference in degree will constitute a difference in kind. We are entering an era of hyperspecialisation – a very different, and not yet widely understood, world of work.

In any given company, hyperspecialisation might reshape the organization in many ways, from the macro to the micro level of task assignment. Some of the tasks of a certain role might be hived off, or entire job categories and processes might be upended. Managers might focus on lower-value-added tasks, as the clients of Samasource do when they hand over data entry. Or they might see greater value in tapping world-class expertise for high-end tasks.

Regardless of task level, capitalising on hyperspecialisation will call for new managerial skills and focus. First, managers will need to learn how best to divide knowledge work into discrete, assignable tasks. Second, specialised workers will have to be recruited and the terms of their contribution settled. Third, the quality of the work must be ensured. And finally, the pieces have to be integrated.

Breaking down the work: Understanding how a knowledge-based job could be transformed by hyperspecialisation begins with mapping the tasks currently done by people holding that job. Such a map may immediately suggest tasks and subtasks that could be performed with higher quality, at greater speed or at lower cost by a specialized resource. In 2008 the pharmaceutical giant Pfizer undertook to do just this in an initiative it called pfizerWorks. Its task-mapping revealed that Pfizer’s most highly skilled knowledge workers were spending 20% to 40% of their time on things like data entry, web research, basic spreadsheet analysis and PowerPoint slides. The company established a process that allowed these tasks to be off-loaded, first to a pair of Indian offshoring firms and then also to an Ohio-based company.

Recruiting workers and assigning tasks: To complete hyperspecialised tasks, companies can use internal employees, develop dedicated relationships with external suppliers or rely on intermediary firms that link clients with communities of specialised workers.

Hyperspecialisation will require most managers to learn to work with the kinds of dedicated intermediaries that have sprung up in recent years to provide access to pools of skilled labour. Much as “cloud computing” services offer on-demand access to computer capacity and storage space, these firms offer “crowd computing” – on-demand access to large groups of appropriately specialised workers.

The intermediaries enable clients to accomplish tasks that range in size from tiny to quite large. On Mechanical Turk and Samasource, workers undertake small tasks that last a few seconds or minutes in exchange for payment ranging from several cents to several dollars. Project sites such as Elance and oDesk enable the completion of medium-size projects in many domains – including web development, graphic design, writing and business analysis – for payments of several hundred to several thousand dollars. InnoCentive and TopCoder undertake complex activities such as software development and scientific discovery for payments that can reach six or even seven figures.

As hyperspecialisation becomes more common, attracting contributions from the most talented workers will become a critical success factor for many businesses. And it will increasingly resemble the way sales and marketing organizations now attract customers: by understanding what people want, figuring out how to give it to them, and learning how to keep them engaged. In fact, cultivating communities of workers is likely to become one of the key disciplines of 21st-century business.


Notify of
Inline Feedbacks
View all comments