Measurable Deliverables? or Miserable Deliverables?

Edward Tsang 2015.01.18; revised 2015.05.23

One way to measure the performance of a person or organisation is to look at "measurable deliverables", e.g. how many goals does a football striker scores? Unfortunately, not everything is measurable; e.g. how much does that striker help the rest of the team? No matter how carefully one defines measurable deliverables, they do not tell the full story of a product or a service. Once measurable deliverables are defined, services that cannot be measured tend to be ignored. When everybody pays attention to nothing but measurable deliverables that partially measure performance, these deliverables could become miserable deliverables.


Measurable Deliverables

How would a manager assess the performance of an employee? One possibility is to look at "measurable deliverables". For example, a football club could measure the performance of a team by the number of trophies that they win. A football manager could measure the performance of a forward by how many goads he/she scores. In a factory, the number of gadgets that the employee produces per hour is measurable. Such deliverables are tangible and measurable. They are objective measures of an employee's performance, which could be very useful.

Unfortunately, contributions may not be measurable

Being able to reduce every person or organisation into a number is great. Unfortunately, not every valuable output is measurable. Counting the goals that a football striker scores is easy. But does he/she create changes for other players? Does he/she help in defence? Does he/she encourage other players? Counting the number of students graduated by a university is easy, but measuring the quality of students is less straight-forward. Counting the number of papers written by a scholar is easy, but measuring the quality of those papers is sometimes difficult.

Does morale matter? If so, is it measurable?

Does morale matter? If it does, then can we design deliverables to measure morale?

It is hard to think of any deliverable that could fully measure the level of morale in an organization. A truely motivated staff would think for the organization, think of ways to improve efficiency, look for opportunities for the organization, etc. It is relatively easy to measure whether a staff achieves goals that he/she is asked to acheive. It is much harder to measure what the staff does not do.

If not careful, measuring deliverables could damage morale. It is hard to measure how much a staff cares. Therefore, the actual cost of implementing measurable deliverables is often unknown to an organization.

Staff tend to deliver nothing but measurable deliverables

Once goals are defined, staff tend to pay attention to those goals. This is because these goals concretely tell them what to deliver. Besides, when staff have limited time and resources, which is often the case in today's management in the name of "efficiency gain", struggling staff would only have the capacity to deliver the bare minimum of what they are asked to do, knowing that when they deliver what they are asked to deliver, they can always defend that they have done their job. Hunger to success would push ambitious staff to produce nothing that is unrelated to measurable deliverables, because they are the only things that matter to their promotion.

Partial measures can easily become the whole

When a manager starts to define measurable deliverables, he/she would typically understand that such deliverables are only partial measures. Unfortunately, such understanding tend to get forgotten over time. Very often, those partial measures become the whole in the end. This will happen especially when both managers and subordinates pay full attention to the measurable deliverables -- they are what everyone discusses about and plans around.

Partial measures becoming the whole, an example

In research, the main goal is creating knowledge. It is useful to write down what one has discovered, because otherwise no one would know. When a paper is written, publishing it is a highly rated journal is useful, because papers in those journals tend to be reviewed more rigorously. Besides, they tend to be read by more people (though this is not necessarily true). The number of papers cite a particular paper might serve as an indication of whether this paper is influential. However, this can be gamed -- a group of authors may cite each others' paper in order to boost their citation rates. When citation is a significant measure, authors will play this game. When the number of papers published is a significant measure, some authors may slice the material of one paper and publish them in multiple papers (these are sometimes referred to as salami publications). Funding helps to facilitate research. However, if the amount of funding is used to measure the performance of a scholar (which is like measuring the performance of a car by the amount of fuel that it consumes), then one could see scholars spending a substantial amount of time to seek funding, which means spending substantially less time on research.

It is rather difficult to define deliverables to accurately measure the amount of knowledge created by scholars. All measures are partial. But if everybody is contended to play the game, it is easy to allow these partial measures to define the activities of scholars.

It gets worse when those who design the deliverables don't have to deliver them

Those who are in a position to define measures for deliverables have power. So even if they have to deliver them, they can always delegate the delivery to someone else. Following the above example, a head of department can spend the department's money to employ researchers to publish under his/her name. Alternatively, he/she can define measures based on his/her own ability to deliver. If he/she publishes in journal X, he/she can define X to be the best journal to publish in. In such situations, the measures are biased. They are further away from measures of true quality.

Measurable deliverables may become miserable deliverables

When partial performance measures become the only measures of performance, activities within an organization or community will change accordingly. In some cases, those measurable deliverable could do a lot of damages to the organization or community. Before the management adopts any measurable deliverables, it would be useful to first examine whether they would distract people from the activities that they want to encourage. For example, in defining measurable deliverables for research, it would be useful to check how scholars such as Isaac Newton and Wittgenstein would perform. If not, then shouldn't we worry about such measures eliminating these scholars from the community?

Example of a miserable deliverables

Suppose the goal is to make sure that a child remembers the times table within 24 hours. This could be achieved through encouragement and inspiring teaching. But it can also be achieved through bribery, threats, physical punishment, etc. This may be effective for some children. They may be able to remember the times table for the exam, which is the measurable deliverable, but there is no guarnatee that they can remember it afterwards. (They are more likely to be able to remember the bribery or punishment.) They may lose interest in education altogether. But interest in education is hard to measurable, and may not be the concern of the teacher who was asked to teach the times table.

[End]

Related: REF as a destructive exercise


All Rights Reserved