The need for a knowledge work productivity management system was identified by the father of management Peter F. Drucker, but in his role he stopped short of providing the details on how to design it. The lack of such a system has been a problem in Enterprises for decades, especially as companies get larger and more complicated.With manual work, the underlying business system is visible. A farming friend in college once explained, at the highest level and in the most simple terms, the underlying system for dairy farming. You feed the cows so the cows can produce milk, and then you sell the milk so that you can keep feeding the cows. With this system it’s clear if the cows have been fed or not, if the cows have been milked or not, and if the milk has been sold or not.In most large companies, unlike the farm, knowledge work is less visible and is difficult to manage because it resides in people’s heads, and there is no underlying system to channel it. We need to be able to manage the invisible, holistic, and ever-changing nature of these efforts in the same way that farmers are able to manage the visible nature of theirs.An important characteristic of a system is that it is a whole with interdependent parts and improving one piece doesn’t necessarily improve the whole unless it is a constraint. In a company, if marketing is not the constraint, investing more money in marketing will not improve the firm’s overall performance. It will actually weaken it by misallocating resources.A system to manage knowledge work requires both a shared framework and an explicit process. First, a shared framework (i.e., a shared mental model) is needed to get everyone on the same page. Then, in conjunction with this shared framework, a standard process is required to help people manage their knowledge work more productively and sustainably.To illustrate the implications of this, I’ll give an example from a business meeting when I was on the Board of Directors of Coca-Cola Nordic Beverages. There was nothing unusual or confidential about this meeting, otherwise I wouldn’t disclose it. But, an example will help shed some light on the nature of knowledge work and illustrate important underlying considerations for making this type of work more productive.Coca-Cola Nordic Beverages was a joint venture between Carlsberg and The Coca-Cola Company, headquartered in Copenhagen, Denmark. During one meeting we discussed whether to proceed with a large multinational technology project. The Board included the CEO, COO, and CFO of Carlsberg, the president of The Coca-Cola Company’s Greater Europe Group, and me.The project we discussed was an important one for the company. One of the Board members was rightly focused on asking What the project was going to achieve. Another member concentrated on Who was going to be responsible for what. A third member focused on How the project was going to be done. And, for a couple of hours, the company’s CEO and CIO fielded a series of questions and follow-up questions and listened to a few personal philosophies and life experiences along the way.Toward the end of the meeting Bill Casey, who oversaw The Coca-Cola Company’s business across seventeen time zones at that time joined the conversation. He shared with the group that 70% of Enterprise Technology projects failed to meet their original objectives and that the average cost overrun was more than 80%. Bill also emphasized that this usually wasn’t because of the technology but was most often due to non-technical factors. With this opening, he asked the company’s CEO to take a couple of minutes to articulate from a corporate view Where he wanted the project to go and Why, What the project needed to achieve by When, How those things could best be done, and Who needed to be responsible for which tasks.Not having had a chance to give a lot of thought to the answer, it was-not surprisingly-challenging for the CEO to articulate this off the top of his head. And, it was even more difficult for the Board to jointly agree. From a knowledge work productivity perspective, this was predictable. It’s rare in important meetings for people to jointly be clear on the Where, Why, What, When, How and Who questions. This occurs with such regularity that people usually don’t even think about it as a knowledge work productivity breakdown. But, that’s exactly what it is.For knowledge work to be managed more productively, as Drucker pointed out, an underlying system is needed. It must get everyone on the same page and properly sequence and accelerate Where-Why-What-When-How-Who. People often are clear on many of these things at an individual level. But, collectively, knowledge workers usually have different individual views that don’t add up to a shared Enterprise picture. In companies and large Enterprise projects, this results in unproductive work and high failure rates. To manage knowledge work better, a system with a shared framework and an explicit process is needed.Initial considerations for the knowledge work systemPeter Drucker wrote that knowledge work needs to be systematized to improve productivity. Manual work productivity, similar to the dairy farm example, can be managed using the objective approaches developed by Frederick Taylor. Knowledge work productivity, on the other hand, requires a combination of subjectivity and objectivity.Using a purely objective approach to manage the fluid and invisible nature of knowledge work has not worked well in practice. When knowledge work is managed like manual work, it tends to get over-engineered, with overly complex governance structures and project designs. Over-engineering work that is invisible, holistic, and ever-changing makes the work take longer and cost more to implement and manage. This explains some of the productivity paradox.Knowledge work productivity often benefits from a “just in time” mindset versus the “just in case” approach that commonly works so well with manual work. With manual work, taking more time to prepare often improves results and reduces risk because the work is stable and won’t change while you’re preparing. With the ever-changing nature of knowledge work, “just in time” is typically more productive and less risky. It often benefits from a prototyping mindset.Prototyping in the field to get work implemented in practice and then making improvements in real time as situations change is often more productive. Knowledge work requires objectivity and subjectivity and an Enterprise ability to discuss, decide upon, implement, and refine decisions better and faster-especially across functions and divisions. Where a good manual work productivity system benefits from being very specialized and mechanized, an effective Enterprise knowledge work productivity system requires a more holistic and better socialized approach.The Enterprise Reinvention knowledge work productivity management system-the framework and the process-requires a minor amount of initial complexity at the front end to avoid an unworkable amount of complication later on. This difference between complexity and complication is more than semantic. Grandmasters in chess, for example, are successful because they apply a certain amount of cognitive complexity up front. By doing this they can view large chunks of the chessboard, whereas amateurs see a mass of individual pieces. In practice, this makes the game much more complicated for less skilled players and makes novices less successful when they play.A key difference between complexity and complication is that complexity has a coherent architecture and can be effectively managed. In contrast, complication is largely random and therefore becomes unmanageable over time. Large Enterprises and large Enterprise projects regularly struggle, not because they are too complex, but because they are too complicated.To illustrate, consider the beverage business. In this industry large soft drink companies are relatively simple from a complexity perspective. They market, sell, merchandise, distribute, and manufacture packaged and fountain beverages. While this work isn’t complex, the soft drink business is actually very complicated. The interactions within and between family bottlers, independent bottlers, publicly-traded bottlers, company bottlers, brand owners, a large and diverse customer base, and a variety of other important parties and personalities make the industry difficult to manage and change.To manage knowledge work, as with the game of chess, it is useful to introduce a minor amount of complexity-a coherent architecture through a shared framework and process-on the front end to eliminate unmanageable complications later. Similar to any good model, it needs to simplify while also being robust enough so that knowledge work tasks can productively self-organize around the architecture in a variety of situations and under various conditions. To build the system that Drucker suggested, a high-level architecture is required, with a shared framework and a repeatable process. The social sciences provide the raw materials to do this successfully, described in the book Reinvent Your Enterprise.