We've decided to make less money: We've slashed our pricing for session replay. They're now more than 50% cheaper for most customers.

Myths (and truths) I've learned from 10 years of startup hiring

Jan 25, 2023

I've spent over a decade hiring for startups, especially early-stage ones. I've managed and hired for teams across engineering, ops, people, marketing, and design, interviewing some 800 people in the process. Some of these teams were highly successful. Others weren't.

These are the destructive myths and important truths I've learned about startup hiring strategy over that time, which I apply to how we hire at PostHog. I hope these tips will help you turn recruitment into a more tolerable and less frustrating experience.

Myth: Recruiters are responsible for hiring

Managers are responsible for hiring. Why? Think about it this way:

  • Managers are responsible for the output of their teams.
  • Output can be increased by hiring more people.
  • Therefore, managers are responsible for hiring.

Recruiters are responsible for maximizing your chances of success by running a great process, helping get a great job ad in the right places, ensuring no balls get dropped, and partnering with you to increase the size of the applicant pool.

If you are struggling to hire for a role, first ask yourself 'am I spending 20% of my time recruiting?' If the answer is yes, then go complain to your recruiter.

Truth: The bar for writing good job ads is depressingly low

There are a million guides on 'writing job ads that wow' out there, and they all seem to miss out the most obvious stuff. We follow this devastatingly simple process:

  • Outline the exact process they will follow (meet X person by name, do Y task etc.). Do not make the process up for each candidate.

  • Write out what they will be working on in specific terms - don't just list generic responsibilities. We find linking to recent GitHub Issues a fantastic way of doing this.

  • Include a salary range. If you can't afford to pay a good salary, then let expensive people self-select out - it's ok!

Congrats, you are now beating 90% of ads out there. Optional extra: add some personality/weirdness.

Recruiters are not copywriters with expertise in every technical detail, so don't force them to write the spec without help. A job spec is an advert for your company, so treat it as such. Recruiting is more leveraged than blog posts, and yet people put in half the effort into the ad copy.

You want people to filter themselves out of the process if the work isn't appealing to them. This will save you time in the long run.

Myth: Hiring gets easier when other companies are making cuts

Not really.

The best people are always in demand and will have lined up five interviews within 24 hours of posting on LinkedIn. Often (but not always) the 'cuts' that are being made are in non-core product areas, or in more manager-y type roles that are unlikely to be relevant to your early-stage startup.

Hiring should always be difficult because your bar should keep going up with each hire. Imagine if developing amazing new product features was easy. It would probably mean that you were building uninspiring stuff, or features that are easily replicable.

Pro tip - reach out to the people who didn't get cut, as they're probably not feeling great about hanging around at those companies any more having seen a bunch of their colleagues leave.

Truth: Getting people to do actual (paid) work for you is 10x more effective than interviews

Interviews are imperfect in so many ways, and should be generally be seen as a reasonably effective filtering mechanism. Getting candidates to do a sustained piece of actual work in a semi-realistic context is much better. Pay them for their time, obviously – it's cheaper than recruiter fees.

This also levels the playing field because a lot of interviews only tell you if this person is good at giving interviews. It's also a great way to sniff out exec-speak BS when you are hiring VP of whatever from fancy massive company – you'll very quickly find out if they can do the work. Don't just make the work 'come up with a strategy' – ask people to actually do the work.

Interviews still provide useful information, you just shouldn't prioritize them at the decision-making stage.

Myth: Your choice of applicant tracking system, aka 'ATS', matters

The most popular tools for startups are similar enough that they won't make a difference between recruiting going well and badly for you. A recruiter blaming your ATS is like an engineer blaming your choice of IDE – it only matters to personal productivity in the very short term.

Don't hire a recruiter whose first priority is migrating to a new ATS.

Truth: Most recruitment metrics are vanity metrics

Number of applications, number of initial phone screens, number of candidates sourced – these are all vanity metrics. It is extremely easy to get 500+ low quality applications for a role – even specialist ones. In marketing, it's the equivalent for optimizing for pageviews and ignoring conversion to paid. Don't do it.

I once managed to get 900 applications in 48 hours for an ops role at a company with pretty low brand recognition that was only advertising on LinkedIn with minimal budget. That doesn't make me a good recruiter.

The number of people at late/final stages of your process, and their conversion rates to offers made are the only numbers that really matter.

Myth: Looking at pipeline data is only something 'bigger' companies should do

For startups, looking at pipeline data is more important and leveraged than it is for larger companies. Why? Because the impact of every minute wasted interviewing and hiring the wrong people is exponentially higher. Look at later stages (i.e. offers, final interviews) to see which channels are most effective for you and be brutally honest about what's working.

For example, we found for a recent SRE role that our conversion from application to interview stage was amazing for sourced candidates. However, 0 out of 6 final stage candidates who were sourced got an offer, which is unusually low – usually our conversion rate at that stage is ~33%. We went inbound-only for that role and the positive impact was immediate. We had fewer applicants overall, but a much higher number of relevant and engaged people.

Further reading