Building a Recruiting Machine

For CEOs of a growth company, recruiting is both the single most important task and the single most time-consuming task; it’s Datavant’s biggest bottleneck and the biggest bottleneck for many companies.

When I reflect on some of the best decisions I’ve made, they’re recruiting decisions.

When I think about what I wish could go faster — recruiting is often the answer. In the last decade, I’ve hired ~500 people, interviewed several thousand people, and spent hundreds of hours more trying to generate candidate interest.

Here are some principles I’ve used for how to recruit:

Part 1: What to Look For

Hire for people, not for roles.

At a high growth company, you should assume that the role you’re hiring someone for will quickly expand, and the needs of the company will be different 12 months from now. As a result, you should be focused on filling your company with the right people, versus over-optimizing on fitting people to the roles you need today.

For non-engineering hires, I use the “GM of Europe” test, asking myself — “if we need a GM of Europe in a year, could this candidate grow into it?”

“Smart, Nice, and Gets Things Done.” Have a simple rubric and make sure everyone knows it.

At Datavant (and at LiveRamp previously), I’ve used an extremely simplistic rubric that the whole company can recite. We want people who are A+ on three dimensions:

1. they’re smart
2. they’re nice
3. they can get things done

This is simplistic to boil hiring down to three traits, but the simplicity of the rubric is what makes is so powerful.

Specific roles may require additional skills to succeed at the role (for example, for engineering roles, you need to be able to code…), but these are the table stakes to be at the company. We don’t compromise on those traits: if someone is brilliant but not nice or can’t get things done, we won’t hire them. Or if we really need a hire ASAP, we still don’t compromise.

Mediocrity is contagious: avoid the pressure to lower the bar.

As companies grow, there are two forms of pressure to lower the bar:

1. The argument that “X isn’t a core role; we just need someone who can execute on the job, not the best in the world.”

I’ve heard this argument for all sorts of roles: Facilities, SDRs, HR, EAs, Marketing, Customer Success, UX Design, Sales Ops, AR/AP, etc. (Engineering and Product are probably the only two functions spared).

For each of these roles, I’ve worked with A-players, and I’ve worked with non-A-players; it’s always possible to find the A-player for the role, and well worth the investment. Moreover, tolerating mediocrity in any role lowers the organization’s standards in general, increasing the odds that the organization as a whole becomes mediocre.

2. The argument that “If we don’t make this hire now, X will happen” (where X is some terrible consequence).

This may be true, and is a real tradeoff. But most of the time, the long-term cost of a bad hire far outweighs the short-term cost of X.

Look for game changers, and hire for strengths.

Every person I’ve ever worked with (including myself) has had major, glaring weaknesses. That’s okay: game-changers are people who have a strong strength that makes them best in the world at something.

You should be specific about what weaknesses are actually disqualifying (for Datavant, not being “smart, nice, gets things done” is disqualifying, but in general other weaknesses are considered manageable).

As long as someone doesn’t have a disqualifying weakness, then you should focus on really understanding their strengths and determining if they can truly be a game-changer for the company on some dimension.

Expand the top of the funnel to improve diversity. Don’t change the bottom of the funnel.

It’s worth the investment to build a diverse and inclusive company. This has to be done in the right way. In addition to striving for a fair hiring process that weeds out bias wherever possible, an important investment is to expand the top of the funnel. Expanding top of the funnel means working harder to get a diverse slate of qualified candidates; it doesn’t involve changing standards at the bottom of the funnel.

Part Two: How to design your recruiting process.

Design your process to minimize false positives. Worry less about false negatives.

A false positive hire (i.e. hiring someone who’s not a good fit) is difficult to correct for. The long term cost of a bad hire is really high. Because firing is so emotionally difficult, most organizations (irrationally) reduce their standards after a candidate joins the company than the standards they have in the hiring process. This means you can be stuck with low-performer for a while. Especially when growing quickly, even a low rate of false positives can add up to a lot of bad hires, and can cascade very quickly into diluting the culture.

Meanwhile, a false negative (i.e. not hiring someone who would have been a good fit) is relatively low cost; it just lengthens the hiring process some.

In other words, if you’re on the fence about hiring someone, bias towards no.

Interviews are terrible ways to assess people, but often the least bad.

Interviews are bizarre situations: they assess how articulately a person answers questions quickly, verbally, and under pressure. For most roles, this scenario is unrealistic.

Unfortunately, in most cases, we’re still stuck with the interview as a way to assess candidates. Here is my rank order of how confident I am in assessing someone:

1. I’ve worked with the candidate directly in the past

2. Someone I trust worked with the candidate directly in the past

3. Projects

4. Interviews

5. Resume

6. Candidate-supplied References

All of these play a part, but design your process to get as close to the top of the list as you can. For example, you should always have a project as part of the hiring process. If the candidate is open to a temp-to-perm role, offer them that before offering a full-time role. Always look for back-channel references. And mine your network, and your company’s network, for referrals from people you trust.

Have a clear process for how you arrive at a hiring decision.

A key part of having a good recruiting process is determining how the final decision gets made.

Unfortunately, this is too often groupthink, which results in people with fewer weaknesses being hired and biases against strong strengths.

On the flipside, a process that is too much in the hands of the hiring manager can result in too many compromise hires.

Here’s the formula we use at Datavant:

1. Each person in the recruiting process indicates whether they are a “Champion,” “Weak Yes,” “Weak No,” or “Veto.”
2. Anyone in the process can veto. If someone vetoes, the candidate is declined.
3. In order to make a hire, at least one person needs to champion — meaning they are going out on a limb to say that the candidate is a game-changer
4. If there are no vetoes and at least one champion, then it’s the hiring managers call.

Part 3: After they’re hired

Don’t decrease your standards for people just because you hire them.
Usually, companies that have a high bar to get into the company lower their standards once someone joins the company. This shouldn’t be the case: managers should always be asking “would I hire this person today given what I know,” and applying the same standards to performance management that they apply to recruiting.

If companies don’t create a strong culture of performance management, the cost of false positives goes up, which makes recruiting even harder.

Keep investing in candidates after you hire them.

Retaining (and growing) a great team member is worth several times more than making an additional hire, but most companies don’t invest that way.

CEOs should obsess over having 0% regrettable attrition rates, ensuring their teams are growing personally as much as possible, ensuring that their compensation for existing employees is always a step above “market,” and ensuring people are fulfilled and likely to stay — all the things they naturally do as part of the new hire process. If a company is doing well and the CEO is paying attention to the team, regrettable attrition should be extremely rare.

PS: Datavant is hiring! Check out open roles at www.datavant.com/careers.