How do you know if you are doing enough at work? How are you doing compared to your peers?
I wish I had an effective way to measure our productivity in knowledge work but this is a very hard problem to solve. Though, as managers, I don’t think we have really tried to figure this out.
The current method I use is to estimate the proper amount of work for an employee based on the amount of work I believe others are doing. If an employee cannot complete this amount of work, they are not doing enough. This is identified by complaints that requests are not being met or by tasks not being done in time.
There are problems with measuring performance this way.
- The same type of tasks can vary significantly in the work required to complete them
- It is hard to differentiate between a superstar who keeps things off my radar and someone with an easy workload
- It is hard to differentiate between someone who has a more difficult workload and someone who sucks at their job.
Some jobs lend themselves to simple metrics. There is an example where salespeople in one company were not required to attend meetings because they had a simple metric to meet and it was understood in the organization that the meetings kept them from meeting that metric. Having simple productivity metrics clarifies the expectations for employees.
Because this is not an easy problem to solve many managers default to hours worked or checking that the employee is at their desk (or checking the little dot on their IM app). This can burn employees out and this can be gamed.
Consider how this is done in manufacturing. The steps in creating a widget are almost completely standardized. We are wired to understand physical objects that we can watch being made but knowledge workers work in their brains leading to unknown process variables. Managers take a hands-off approach because they cannot see the work being done and put the expectation on the employees to improve their productivity. We give them tactics like “check your email less often” or “block out time in your calendar” expecting this to improve productivity (this is like telling a machinist to make their machine go faster without the maintenance resources to make modifications and continuing to give them work by piling it on their workbench).
To better measure knowledge-work productivity we need to standardize and improve processes (like using a work management tool rather than email) and reduce the variables in completing tasks.