Here's a brainteaser: You are given a candle, a box of tacks and a book of matches. How do you attach the candle to a wall so that it can be lit without dripping wax onto the floor below?
The solution to Duncker's candle problem is to deconstruct the box of tacks into its parts (box, tacks), attach the box to the wall with the tacks and attach the candle to the bottom of the box. In experiments, people who get the tacks inside the box can't solve the problem, but those given a pile of tacks beside the box solve it easily.
What does this have to do with HR? Typically, work is “constructed" into job descriptions similar to the box of tacks; the descriptions become a repository of competencies, performance indicators and reward packages. The result is that powerful opportunities to optimize your workforce's productivity, alignment and engagement are often obscured. Tapping into those opportunities requires “deconstruction" and “reconstruction," a vital emerging capability within HR.
In the dynamic work environments of the future, HR dilemmas closely resemble the candle puzzle. It's time to take the tacks out of the box.

The Hollywood Model of Work

Job descriptions usually describe work performed by full-time employees, throwing together a series of skills and requirements that seem large enough for an employment contract. But as my co-authors (Ravin Jesuthasan and David Creelman) and I suggest in “Lead the Work," regular full-time employment is only one way to get work done.
Roger Martin argues that "knowledge work" never belonged in jobs, and that companies mistakenly applied an industrial age work model to knowledge work; instead, knowledge work is better organized into projects. If you deconstruct work into projects, you can assemble teams of employees and free agents, like the Hollywood Model of movies and television. 
Free agent platforms like topcoderTongal and Upwork deconstruct jobs or assignments into parts that have a tangible result. Here, free agents become an option because there is less need to monitor skills, effort, diligence or activity, as long as the assignment is completed at an acceptable quality and on time. Such platforms thrive on work such as data compression algorithms, internet or TV commercials, and entering figures from receipts.

Finding the Real Performance Indicators

With set job descriptions, it can be difficult to break down the “return on improved performance" (ROIP) of work. For example, sweepers at Disneyland both sweep the park and help guests, but they usually make the biggest impact through their great interactions with guests (as long as the sweeping is good enough). However, typical job descriptions would list both sweeping and guest interaction as key performance indicators.
As another example, Boeing Co. aerospace engineers are tasked with designing aircrafts and collaborating with suppliers to deliver them. But suppliers often have more technical knowledge about composite materials than the Boeing engineers, making supplier collaboration a more pivotal part of the role than a job description might suggest.
What happens if you don't deconstruct the job descriptions? Sweepers may mistakenly strive for excellent sweeping over guest engagement, and engineers may mistakenly strive for excellent technical design over collaboration—because “it's all part of the job."

When Everything is Important, Nothing is

In jobs like "science director" in the pharmaceutical industry, "software architects" in technology or "chief science architect" at a non-profit, the descriptions often include everything from publishing scholarly papers, to generating patents, to running processes (like drug development, software development-testing and charitable fund allocation).
Leaders in these jobs often believe that every work element is pivotal, but individuals who are excellent at everything are rare. So, these leaders are often frustrated, because they are admonished for the job elements where they are “only good enough," yet they know they are excellent at other elements.
HR technology systems reinforce this by screening for job candidates that fulfill every qualification (such as possessing a Ph.D. and a lengthy research vita plus years of experience leading technical teams, plus global fame as a thought leader). This causes frequent failure to produce ready candidates on time.
Deconstruction reveals a more nuanced reality: The same job adds very different value in different situations. Sometimes, “leading the team" is most pivotal and “scholarly research" can be good enough. Other times, it is the opposite. Deconstructing reveals the “return on improved performance" for each part of the job, illuminating new solutions.
What can you do now? Think about the “work," not just the “job." Encourage your leaders and key jobholders to take those lengthy job descriptions, deconstruct them into their parts, map the ROIP of each part and put them back together in new ways.