I’ve worked at startups from 4 people to 400+ people and invested and advised from incubation to Series D and beyond and it still surprises me how often “how do we hire?” comes up as a topic of discussion. I mean that both in the macro sense (“how do we know if we need to add a new person”) and the micro (“how do we effectively vet people to minimize false positives and false negatives?”).
This guide covers that whole spectrum, though possible not at the most abstract. While I won’t be tackling “when should I add a new person”, if you start with “how do we define the role” it can be a helpful inversion: if you can’t define the role, you’re certainly not ready to hire for it.
I’m mostly focusing here on hiring for startups after they have some money in the bank and likely >25 people (approximately). It’s a “paint-by-numbers” guide to kickstart how you think through hiring that can establish good hiring hygiene. Remember to not get too bogged down in process: process is only value accretive inasmuch as it helps you move faster; too many people like process for its own sake, which is a trap. Documenting this full process is helpful in either “low-trust” environments (it’s a person’s first time hiring) but also in distributed or remote environments, where written communication is much more important to team cohesion.
The first few times you do this it might take a couple of iterations to get right, but once it’s ingrained in how you think it should take no more than an hour to complete for a role you’re familiar with. Naturally, it’ll take much longer if you’re hiring for a role outside your area of expertise, where you don’t even know what you’re looking for let alone how to vet people. (In that situation, the time investment upfront will save you a lot of anguish later.)
With all of that out of the way, let’s get started.
If we add the ideal version of this person, what will be true of the business in 30 days, 90 days, and 180+ days? What are the outcomes we need them to achieve over each of those time horizons? What problems do we need them to solve?
Focus on outcomes, not inputs. This goes both for your internal and external (read: the public job description) language about a role. You don’t have to use the exact same language (some things might be sensitive and for internal consumption only) but focus on outcomes in both cases. A given role will have multiple outcomes it needs to achieve over varying time horizons.
Depending on the level of the role and the stage of development of the startup, the scope and detail of these outcomes can vary wildly. For example, good framings of outcomes:
Bad framings look like:
Be rigid on the outcomes but flexible on the methods. Why? Because this both focuses on the right things — the impact on the business — and communicates autonomy to candidates. It says: “we trust you to do things the most effective way possible”. It doesn’t mean you abdicate oversight or responsibility for their actions, but it does say “we are a team that values leverage and action above process and box-checking”.
It could be that you have strong priors on what the right answer is in some situations, but that doesn’t need to go in the job description. For example, it might be right that the executive team needs a weekly report on CAC by channel, but reading that in a job description doesn’t get anyone excited about a role, and it if it’s in the internal language about the job it sends the subtle message that this person is an order-taker, which is the last thing you want in the culture at a startup.
Other things you need to align on internally, that might need to be reflected in the language of the job description:
There are 3 main tools people use to assess candidates. I’ll list them in descending order of information quality:
If you get a screaming hot referral for a role from someone you trust intimately you can accelerate a lot of the interview cycle to get to an answer. But remember to always trust but verify. You still need to ensure the candidate can do what you need (which is why defining the role comes first before doing any outreach) and that the candidate wants to do what you need. And trust yourself first: you know what your business needs, so even if it’s a high priority referral if it doesn’t feel right don’t rush to hire.
You need to codify what you’re looking for in a role into a rubric. A rubric might look something like this:
Next, distill that rubric into vetting methods to assess each element. Some of this might mean pre-writing interview questions to be used on all candidates for the role, but it goes beyond questions: some rubric elements might really demand a work sample, case study, or presentation to your team to assess effectively. Others might only be revealed in reference checks to get a longitudinal understanding of a person’s behaviors. Don’t assume everything can be assessed in an interview. Remember to also sketch examples of what good or bad responses to different vetting methods might look like, so the hiring team is aligned on what a good candidate actually behaves like.
(I’ll write separately about how to conduct a good reference check and how to design an effective case study.)
It’s easy to design a rubric (and vetting methods) for a role you’ve hired successfully before or one within your domain of expertise. It’s much harder to do so for anything outside of that. For example: if you’ve never hired lead data scientist before, and that’s your first data science hire ever, you may not know what an effective lead data scientist looks like. You also won’t know what some common false positives are.
To solve this, call up people you respect in the domain you’re hiring within to get their help. Describe to them the business outcomes you want to achieve with this hire and ask them what they would look for, what they would avoid, and how they would vet candidates. If necessary, pay them a consulting fee to help screen or to conduct interviews for you, especially if proper vetting requires asking technical questions that you’re ill-equipped to assess.
Once you have the rubric and vetting methods defined, design the logistics of your hiring funnel. Literally: what’s the order of operations and who is responsible for which part? Don’t forget to build in space for the hiring team to sell the candidate on the role and the company, and for candidates to ask the hiring team questions. This isn’t a one-way street.
Finally: be flexible how you approach hiring. Depending on the company, the role, and the market, sometimes you can put a take-home skills assessment early on in the funnel. Other times you’ll need to do a lot more selling before a candidate is excited enough to put that work in.
The best option, as mentioned above, is warm referrals. They can come from employees, friends of the company, investors, customers, and more. And remember, trust but verify.
Beyond that, if you’re doing cold outreach, focus on companies that have faced the challenges you’re going through to find the most likely analogous candidates. They might have different job titles, so be sure to prioritize the work being done (and the competencies that indicates about a candidate) above an exact title match.
Sourcing is difficult partially because the default strategy is undifferentiated: cold outreach, blasting away at people’s LinkedIn or email inboxes. One of the best longer-term hiring investments you can make is in your company’s brand as an employer. If you can show potential investors that you’re a talent magnet and attract exceptional people (especially by your Series B), that is a strong signal of a high quality company.
That’s plenty to get started with, and I’m sure there are parts I’ve left out or got wrong, so feel free to respond with comments, suggestions, and additional advice!