Hiring is a matching problem


Hiring is a matching problem: finding the right fit (the right employee + the right company) is hard but vital. Investing more effort into this search will pay big dividends.

What is a matching problem, and why are they hard?

Everyone has first-hand experience with matching problems: dating. When you first meet someone, you hope they are fantastic and that you'll spend the rest of your life with them, but you're not sure. Moreover, you can't be sure – no amount of Tinder browsing or hearing rave reviews from friends can replace actual experience with the person.

Your problem: figuring out whether the two of you are a good match. Solving this is vital. Effort spent on finding the right person will lead to much better results than trying to salvage a doomed partnership.

This applies beyond just the world of romance. In many domains, the improvement from finding a good match far outweighs the amount of work done afterwards. Think organ transplants, blood transfusions, university dormitory placements, and employment.

The trouble is, matching problems are hard. Economist Al Roth won the Nobel Prize for his work on the topic![1] They are hard because of information asymmetry: each party knows something the other side doesn't. Compounding this, neither party can credibly signal upfront that they have something the other side wants. Anyone can claim they share your love for arthouse movies or passion for beanie babies – the only way to verify is to check for yourself.

Which means that matching is costly. But it is far more costly to ignore the problem, and just run with whatever you pick first. By doing so, you are optimising locally rather than globally; climbing to the top of the hill is laudable, but before you invest all that effort you should look around and make sure that it's the right hill for you!

Employment is a matching problem

Employment is a matching problem: a company wants employees with certain skills and the right attitude, and the employee wants a company where they will thrive. Both sides can claim that this will be the case, but the truth of the matter will only emerge over time, as the employment relationship unfolds. This is both because the definitions of ‘right attitude’ and ‘place where I will thrive’ might vary in important ways – ways that won’t be obvious until you roadtest the working relationship.

Best case scenario: everything is great, and the employee thrives while adding enormous value to the company. More likely scenario, especially early in an employee's career: the fit isn't quite right, and neither side is as happy as they hoped. This usually isn’t due to bad intentions or either side lying. It’s more that you learn what you need by experiencing it (both positive and negative examples), so someone early in their career may not either know or be able to articulate exactly what is important to them.

And a bad match that isn’t abandoned will be costly!

Let’s say that a good match between the company and employee leads to the employee being 10% more productive, with these gains compounding over time. After just five years, the employee who matches well with their company will be 61% more productive. And after twenty happy years with the company, this employee is 670% more productive![2]

From the company’s side, there is also the risk to the culture. An employee who fits in perfectly will reinforce and spread your culture much more strongly than an employee who isn’t the right match.

Why might the match not work?

One reason is a lack of cultural fit. Say, the employee needs a very structured work environment, but the company is very hands-off. Or the company expects staff to be available round the clock if something goes wrong, while the employee values strong work/life boundaries.

Another key reason is a mismatch in responsibilities vs the sort of work the employee is best suited to. Brad is a people person stuck in a job where he works alone most of the time. Or Elise likes data but is required to do more qualitative story-telling. Or it could be at a more granular level: maybe design work is a good fit, but your passion is for branding rather than UX/UI topics.

What should we do about it?

Companies and employees should spend much more effort finding the right match. A diligent employee is going to work hard regardless, and a smart company is still going to try to give them what they need, but if the match isn't right, both will be running uphill.

Loyalty is a noble and important sentiment, but risks a colossal waste of resources from both sides.

So how do we do this?

For starters, in the interview process, the company should concentrate on identifying whether the candidate could thrive for many years. Same for the candidate. They should use the interview to identify whether the company will provide the conditions critical for their satisfaction and productivity.

After the candidate is hired, the first 6–12 months should be an evaluation period for both sides. The critical question to ask is: "Did we make the right choice?" Compensation and incentives should be structured to reflect this, for example:

  • 12-month clawback period on upfront bonuses
  • 1-year cliff for vesting
  • Potentially back-weighted vesting

Lastly, the matching problem should be the top concern for people managers with new employees. Every week, ask yourself whether your new direct report is performing and fitting in as well as you had hoped. And you should regularly ask them whether the job is what they expected, and whether it is working for them.[3]

Final thoughts

Lots of good employees have spent time early in their careers searching for a fit before they found a good match with a company and settled in for the long haul. Take a poll of your colleagues: I bet you find many who have plenty of sub-two-year job tenures at the bottom of their CVs.

Companies and employees should invest more effort finding a good match, because the payoffs will be enormous from doing so.

Additional resources

Wait But Why has some great advice for people at the start of their careers who are facing the matching problem: https://waitbutwhy.com/2018/04/picking-career.html

Josh Kaufman explains the explore/exploit technique from data science as a way to solve the local optimisation problem in job search: https://joshkaufman.net/explore-exploit/

[1] https://www.wikiwand.com/en/Alvin_E._Roth

[2] This model is certainly a vast over-simplification and makes a big assumption about whether productivity compounds endlessly. Balancing this, I suspect that 10% is an enormous understatement for the gain from finding a job you love. See http://paulgraham.com/love.html for some reasons why the real increase from a good match is probably much higher.

[3] Your 1:1 meetings (which you are having at least fortnightly, right?) will be perfect for this.