An engineer’s take on improving productivity of knowledge workers: Part 1 (Cartesian method)

How to sustainably improve the productivity of knowledge workers is one of the foundational unsolved problems in People Analytics. I’ll admit upfront that I don’t have a ‘theory of everything’ that solves this problem conclusively. I am writing this blog in part to clarify my own thinking (probably very skewed by my engineering background) and to get input from others who may help me build on it or offer a different way to look at it.

We all know that Ford’s assembly line improved the productivity of the car manufacturer 8 fold. While this is truly impressive, there are physical limits to which manufacturing productivity can be improved. There are no such limits to improving productivity of knowledge work. Moreover, I’d argue that the knowledge economy represents majority of the GDP today. How might we take lessons from Ford’s success to the knowledge work? I’d say that Ford applied 3 problem solving techniques to improve productivity

1. Cartesian method: Divide a complex problem into components which are less complex and repeat until the parts become simple to solve.
2. Systems thinking: Understand how different components relate to each other i.e. dependencies, relative importance, relative effort
3. Math: This may feel like a stretch. But the final design of how work should be done is essentially an equation that helps the different components add up into the final product.

In this blog I’ll explore the application of Cartesian method (sometimes referred to as Issue tree approach) to the productivity problem. We’ll look at how it can help address productivity challenges in response to events like COVID-19 + Working from home (WFH)

One way to break down the productivity problem could be

First two levels of the productivity issue tree

1. Solving the right problem: This is the most important part of productivity in the knowledge space. If we are solving the wrong problem then nothing else matters. Executives at the company should shoulder the majority of this burden. They need to develop the right strategy and ensure that everyone on the team is pulling in the same direction. There needs to be a two way feedback to help correct the strategy if needed.
2. Having the right people in place: Even if we are solving the right problem we need the right talent in the right roles to solve the problem. We can also argue that we need right people to ensure that we are solving the right problem. At a high level this can be broken down into acquiring, nurturing and motivating the right talent.
3. Having the right resources available: To solve the problem successfully the talent needs to be equipped with the right resources. They need technology, productive time and money/investment.
4. Taking the right approach: The makeup of the team, how they work together and the culture of the team has a large impact on productivity as well.

At this point we have broken the problem down into pseudo-MECE (mutually exclusive and collectively exhaustive) components. But this level of granularity is not very useful to solving the problem. The components are too generic and at the same time each component is complex and ambiguous. So let’s double click one more level

First 3 levels of the productivity issue tree

At this level of granularity the issue tree has a little more practical use. Let’s go back to the beginning of the pandemic. Every CEO was first and foremost concerned about (hopefully) the safety and well being of their associates and customers. But close second was the concern about the effect of COVID + WFH on productivity. The issue tree above could be used to quickly evaluate which components may experience stress due to COVID + WFH. This can help leaders prioritize their efforts to maintain productivity.

Components that may experience more stress due to COVID + WFH

If I were the CEO of an organization I’d hypothesize that the components highlighted in Red are more likely to experience stress due to COVID + WFH. The benefit of this approach is that it helps leaders make their logic explicit and makes it easier for others to provide their input before a decision has to be made. As a leader there is never any excuse of having bad logic or not making their logic explicit.

Beyond logic there can be a lot of value in precision of insights e.g. the extent to which we can move the needle on one component, the level of investment it would need and the level of impact it would have at the overall level . The level of precision needed depends on the opportunity cost of being wrong. We can keep going deeper on the issue tree to get more precise. Going two levels deeper can make the components far more actionable. But if we just use the Cartesian method then we risk getting overwhelmed because the number of components increases exponentially as we go deeper. When we are overwhelmed it becomes very tempting to start throwing out ideas (spaghettis on the wall) rather than working through a problem systematically. In the next two blogs I’ll explore how Systems thinking and Math can save us from the temptation of spaghettis on the wall approach.