The 3 rules of measuring success

How the metrics used to measure success shape you and your future

The 3 rules of measuring success are:

  1. Metrics used to measure success apply evolutionary pressure on people, traits and behaviors. E.g., if you are a writer and you measure success in terms of clicks, not only will you become better at click baiting and you will edify your skills at writing original, valuable content, but you will also select your readers based on their susceptibility to clicking rather than on their propensity to pay you for your content.
  2. Unless the people, traits, and behaviors needed for short and long term success are the same, choosing a short-term metric will cause you to plateau after the short term, whereas adopting a long-term metric will cause you to plateau only after the long term. The short-term metric will cause you to develop traits, behaviors (and customers, friends or employees, depending on your context) which are advantageous for achieving your short-term goals, but detrimental to achieving your long-term ones. Corollary: success unsupported by working skills or achieved through shortcuts is almost always followed by failures.
  3. Choosing a metric because of an external imposition, suggestion or expectation will lead to regret and unfulfillment. Corollary: the fundamental failing in human decisions is optimizing for a metric just because it’s possible to do so.

Let’s examine the three rules in greater detail.

Rule #1: Metrics used to measure success apply evolutionary pressure on people, traits and behaviors.

If a behavior is rewarded, people will adopt more; if it isn’t, people will adopt it less. This is fairly obvious. The actual process to which this happens, and its consequences, are not. To better understand it, it is useful to highlight the similarities between this process and Darwinian evolution.

In Darwinian evolution, animals develop random mutations. If a mutation provides an increase towards the success metric (fitness: the animal is more able to survive in the produce), that mutation will get adopted by more animals. If it does not provide any increase, that mutation will disappear instead. The result is that animals end up adopting those mutations which provide advantages towards the success metric enforced by nature and end up rejecting those which do not. Humans do the same with behaviors. If a person exhibits a behavior and is awarded because of it, she will tend to produce it in the future; if punished, she will not repeat it. (Notably, if a person exhibits an effortful behavior and is not rewarded for it, she will abandon it. A costly unrewarded behavior is indistinguishable from a failure, in the eyes of its performer.)

Which behaviors thrive in a culture is a function solely of the rewards or punishments they attract, and it is the rule which determines whether a behavior is rewarded or punished that forms the success metric used by individuals to make the decisions. Therefore, adopting a success metric places an evolutionary pressure on behaviors, determining which ones are developed and which ones disappear.

Let’s dissect the previous statement into three parts:

  • A success metric promotes those behaviors which provide an advantage towards them.
  • Similarly, a success metric discourages those behaviors which do not provide an advantage towards it. This is worth remembering for two reasons. First, behaviors which are neither rewarded neither punished are not left untouched: they are hindered instead (because every behavior has a cost, and cost + no reward = negative result). Second, forgetting to reward the behaviors which are not good for short-term objectives but are good for long-term ones will cause them to disappear, ultimately making the long-term goals unattainable. This point will be further explored in rule #2.
  • A success metric is to be considered “adopted” if and only if rewards and punishments happen based on it. A success metric which does not result in rewards or punishments is invisible to our intuitive mind (which is the one taking decisions). Moreover, we humans are excellent at inferring a success metric given a set of rewards and punishments, even if no formal explicit success metric is present.

The personal interpretation/inference of the rules which predict when subjectively perceived rewards and punishments are administered determines which success metric is used by individuals to take decisions, and later mine which utility functions are to be optimized with their actions.

As Ben Thompson wrote[1] “culture is not something that begets success, rather, it is a product of it.” Culture does not create success metrics; rather, it is a product of them.

Superstition

Success metrics which are not entirely aligned with our long-term goals promote undesired behaviors by introducing unwanted rewards to wasteful actions. This complex phenomenon is commonly referred to as “superstition”.

People adopt superstitious behaviors after receiving an uncorrelated reward or punishment immediately following an action of theirs. Given the immediacy of the reward or punishment, people assume correlation, or even causation, and accordingly repeat or avoid their action in the future.

We all know how stupid it is to assume that “a T-shirt is lucky” just because we were wearing it the day we won the lottery. And yet, many of us assume that a shirt fits well on us just because our date kissed us the day we wore it. Similarly, a parent might believe that punishing his son after a bad exam has been a good idea just because the next exam goes better, even though the variation of the marks is entirely explained by random statistical variation and by regression to the mean. Likewise, a boss might assume that his aggressive behavior is the cause of the good performance of his team, simply because just after the first time he got angry his team performed well, for completely unrelated reasons (perhaps, by chance). I would go as far as to say that most bad behaviors which are exhibited by only a percentage of the population are solely the result of superstition: people got rewarded for it in the past and they repeat them ever since remembering that time. They learned, in their experience, that such behaviors lead to positive results.

Accepting a success metric which is not entirely aligned with one’s long-term personal goals leads to the development of superstitious behaviors: behaviors which are rewarded by a success metric which is not the one that should matter, and are thus reinforced even if detrimental to the success we ultimately need.

An example: let’s say that we decided to adopt money as a success metric of ours. If a behavior of ours causes us to make money, it will get a reinforced, even if detrimental to our long-term happiness. This is why people overwork themselves to the point of neglecting their families or their health.

I particularly like the formulation of “success metrics apply evolutionary pressure on behaviors” because of the following analogy. Naval Ravikant said: “A cockroach is just as evolved as we are, just across different fitness functions.” Similarly, every person is just as evolved as everyone else, just across different success metrics. Everyone’s behaviors are just as rational[2] as everyone else’s; simply, we all optimize our actions to follow different utility functions. Personal evolution is inevitable; by choosing which success metrics you will allow yourself to be rewarded by, you will shape yourself. Don’t let yourself become a cockroach.

Rule #2: choosing a short-term metric will cause to plateau after the short term

As per rule #1, adopting a short-term metric will promote behaviors advantageous to reach the short-term goal and will discourage those which do not provide any advantage towards it. In particular, adopting a short-term medic will cause the people and behaviors which are beneficial for the long-term goals to disappear (if they are not also beneficial to the short-term ones). Once the short-term goals are reached and attention shifts towards the long-term ones, the behaviors needed for success towards such long-term goals will not be there anymore. Let’s make two examples to clarify.

A student who decides to adopt the short-term goal of passing his exams will find that memorizing the textbooks tends to be more efficient and effective than actually understanding the concepts written in them. Therefore, during his university years, he might develop exceptional memorization skills but no real understanding of the topics of his degrees. Such memorization skills will help him to pass all his exams without spending too many hours on the books or at the lectures. However, once he finds a job, and shifts his focus from passing exams to getting promotions, he will realize that he lacks the conceptual understanding needed for his new goal. What helped him reach his short-term goal (his memorization skills) caused him to neglect the skills he would need for his long-term goal (a successful career).

An online publication might decide that its short-term goal is to have as many people as possible to read its article. Its strategy is that only once it will have reached a given number of readers, it will try to convert them to paying subscribers. In the first months, the success metric it will adopt will be the number of clicks journalists get for their articles. Such success metric will reward behaviors such as click-baiting; journalists will be rewarded for writing catchy titles. Those writers whose skill sets are about writing original content rather than attractive titles will find themselves unrewarded and will leave the publication. After a few months, the publication will try to convert its readers to paying subscribers. It will encounter a problem: since it selected its readers based on their willingness to click on sensational titles rather than on their willingness to pay for valuable content, very few of them will subscribe. Realizing their mistake, the editors will ask the journalists to write more valuable content. There will be a second problem: the journalists able to write it already left the company, and those still there have the wrong skill set (writing catch titles).

Both examples are expressions of rule#2: choosing a short-term metric will cause to plateau after the short-term. The student who chose a short-term metric got his short-term success and then plateaued afterward, as he lacked the skill set needed for the long-term success. The company which chose a short-term metric got its short-term success and then plateaued afterward, as it lacked employees with the skill set needed for the long-term success.

In both cases, short-term goals were decided with the assumption of living in an ideal world deprived of second-order effects. Rewarding a behavior means not rewarding another one; promoting a behavior means discouraging another one; welcoming people with one characteristic means leaving those with an opposite one.

Moreover, another assumption was made: a static world, made of linear progressions, which require the same skill set. This is seldom true; more often, a rule is valid: what got you here won’t get you there. Different stages in life require different behaviors to excel at them. For example, the behaviors that one has to display to be considered a good student are different from the ones that have to be exhibited to be hired, which are distinct from the ones that have to be the shown to be called a good employee, which are different from the ones that have to be displayed to get a promotion to middle management, which are distinct from the ones that have to be mastered to be raised to the executive rank. Optimizing too much for a stage causes to develop a skill set which is an adequate for the next stage. Instead, keeping an eye on the long-term leads to developing a skill set that is as common as possible to all stages, and that allows to produce the behaviors required by each one, perhaps less efficiently but more consistently.

Rule #2 is valid only in those contexts where the behaviors required for short-term success are different from the ones required for long-term success. For example, a company might find that the short-term goal might be good to focus and energize its employees, as long as the behaviors it fosters are also beneficial to its long-term goals, and as long as doing so will not cause its employees to atrophy their behaviors which are good for long-term goals but have no influence over short-term ones.

Rule #3: Choosing a metric because of an external imposition, suggestion or expectation will lead to regret and unfulfillmenT

“Would my family think it is admissible for me to pursue this goal?” “What goal are others expecting me to pursue?” These are just some of the questions that pop into our minds when deciding which goal to set for ourselves. Every time we let one of these external pressures to influence our decisions on which goal or which success metric to follow, we set ourselves up for regret and unfulfillment.

Some examples:

  • Unfulfillment arises from using metrics chosen by others as proxies for personal success (e.g., a new car as a proxy for happiness).
  • Failed relationships arise from choosing a person using widely-desirable traits as a proxy for personally-desirable ones.

The reason for such regret and unfulfillment is that our subconscious is the part of our mind that will ultimately judge whether we are happy or not, and it will make such judgments based on its own criteria. Using someone else’s criteria for deciding what to do is stupid: unless those criteria are identical to our intuitive unconscious (they never are), there will be discrepancies in the emotional results of the actions taken, and this will lead to satisfaction. It is like asking your mother for what she likes eating and then the cooking that for your girlfriend. What a surprise she didn’t like it! Doing something for someone to satisfy someone else’s desires is bound to leave the first someone unhappy (that would be your intuitive subconscious).

A consequence of rule #3 is that you should never adopt a metric just because you can. Inevitably, adopting a metric will cause you to do something you would not have done otherwise (or it will cause you not to do something you would have done otherwise). If there is a way you could maximize the secondary success metric without contributing to your main one, it will happen. Such a waste of energy, focus, and resources. Moreover, unless the success metric that causes you to act in that way is perfectly aligned with your heart’s desires, your heart will not like the way you acted. This concept is simple, and yet frequently forgotten. Intelligence is the ability to optimize over a given metric. Too often, we are so eager to show our intelligence that we lack the wisdom to ask ourselves: is that metric worth being optimized?

Please use the clap button at the side to help other people discovered his story.


Notes:

[1] “The Curse of Culture”, stratechery.com, May 24, 2016

[2] The definition of rational is not logic, but maximizing the utility metric considered.