You Are What You Eat

A few stories about biases and their effects.

East Campus

At my alma mater, every year, each dorm holds rush activities for incoming freshmen. It’s similar to other recruiting functions—have events and activities that draw people to your dorm.

My dorm was an interesting one. It was an old building, with student-painted murals covering nearly every square foot of hallway. The student culture was strong and strange—it was countercultural, weird, and loud.

There were many subcultures, with a wide variety of traditions, characters, and favorite activities. Some drink hot cocoa on a nightly basis. Some identify as loving math. Some light things on fire for recreation.

Now, during freshman rush, there was a rule—don’t talk about singles and don’t talk about price. Our dorm consisted of mostly single rooms, unlike many other dorms on campus. It was also in the cheapest tier of housing prices.

The rationale was simple. We don’t want people here because they want cheaper housing or a single room. We want people here because they want to be here. By advertising culture instead of price or space, we influence incoming freshmen to self-select to our values.

Advertising Perks

It’s common for job advertisements to focus on perks, especially in tech. People talk about their snack selection, office space, amenities, etc. To some degree, this is price of admission. It’s generic advertising, but companies feel that they have to advertise these things to seem competitive.

A company I follow likes to promote its rooftop views. Getting to soak in the sun while looking out over the city…at work? Who wouldn’t want this?

However, this self-selection doesn’t make sense to me.

First, I don’t think there’s a positive correlation between people who enjoy rooftops and people who do well at their jobs.

If the variables are independent, there seems to be three good cases. One, someone on the fence uses your perk as a tiebreaker over their other job offers. Two, you increase your top of funnel and rely on your ability to differentiate quality to higher more good people. Three, you are aiming to rapidly grow headcount and are willing to lower your bar a bit in exchange for speed.

In the worst cases, there’s some kind of negative correlation with job performance. Maybe enjoying perks comes at the cost of skill. Maybe the demographic that would respond to your perk ads are not as strong of a demographic. Regardless of why the correlation exists, this adds a burden to your hiring operations and hires people who perform worse.

For what it’s worth, I think the reality is somewhere in the middle—perk ads offer negligible value while requiring some kind of upkeep.

Free Market Thinking

In some tech organizations, management seems to think of recruiting and retention primarily in fiscal terms.

It is viewed as a transactional relationship—if the company and the employee both feel that the employee’s compensation is fair and competitive, then the employee should be content. If the employee is unhappy, it’s a free market.

Relatedly, management views compensation as a lever. If the company needs to deal with some kind of recruiting or retention difficulty, it can pay more to alleviate the problem. Problems might include a competitive market, low morale, growth objectives, etc. To an extent, this works.

People, of course, don’t work that way. They can be emotional, subjective, and irrational. Different people assign different values and priorities to different things. Even if someone agrees that their compensation is fair and competitive overall, they may still be upset due to their mindset, the culture, and current events.

The failure mode is a downward, self-reinforcing spiral. This happens when there’s some prolonged difficulty that affects a significant number of people. This has been highlighted during the pandemic, but the failure mode has always been there.

As a hypothetical, let’s say the challenge is that a company has a lot of unsatisfying, tedious work that has to get done. That could play out something like this:

Employees are upset because their work is unsatisfying and taxing. A few people decide to leave because they want work that has more meaning to them.

This work still has to be done, so someone(s) end up picking the slack. However, the nature of the work hasn’t changed—it is still unsatisfying and taxing. If some of these people also decide to leave, we now have a recurring problem.

Management notices, and decides to try to fix the issue. But management seemingly only has a hammer in their toolbox, and that hammer is compensation. They decide to raise salaries, give bonuses, or increase equity.

This has a short term effect of making people happy. The work is still undesirable, but at least you’re getting paid to put up with it. However, people will eventually normalize their expectations and become unhappy again. Then they leave again.

This is self-reinforcing. Management continues to increase compensation to address the problem while employees continue to be unsatisfied at work and leave. By default, this is a downward spiral. If left to reach a critical threshhold, you suddenly have a huge problem on your hands.1


  1. For the curious, I think these should be resolved by addressing the problem directly or by using short-term levers to buy time for long-term levers to take effect. In this scenario, this could mean changing the nature of the work (e.g. automation) or finding and hiring people who enjoy that kind of work. In general, this requires recognizing that you are biased or blinded in how you are approaching a problem. [return]