On hiring slowly

The single thing the studio has done right, on the hiring axis, is to hire slowly.

The studio has made every common small-studio hiring mistake at some point. Hiring on personality fit and not on craft. Hiring on craft and not on personality fit. Hiring before the workload existed to keep the new person occupied. Hiring after the workload had become so heavy that the new person was thrown into a fire on day one. Hiring without a clear definition of what the role was. Hiring with a definition that was wrong for what the studio actually needed.

A few rules the studio has settled on, after enough mistakes.

The role does not exist until the work has been turned away for three months. If the studio has been turning down projects, or pushing back on timelines, or staying late on existing projects, for at least a quarter, the role might be real. If the workload is variable, or the studio has been complaining about being busy for a few weeks, the role is not yet real. Hire only on sustained pressure, not on temporary spikes.

The role description is written before the candidate is identified, not after. Most small-studio hires happen because someone the studio knows is available. The role is then shaped to the person. This works occasionally. It fails more often. The role written for a specific available person tends to drift from what the studio actually needed once the person joins. Write the role first. Compare candidates against it.

A two-week paid trial, before the full-time offer. Hand them a small project. Real work, properly scoped, properly briefed. Pay them at their day rate. Both sides learn whether the fit is real, in conditions that are closer to the actual job than any interview process. The studio has, on more than one occasion, parted ways at the end of the trial without bad feeling. Both parties had information they would not have had after a normal hiring process.

The trial concludes with a written debrief from both sides. The candidate writes what they observed about the studio. The studio writes what it observed about the candidate. The two documents are exchanged. The conversation that follows is more honest than any interview, because the work has already happened.

The first three months are protected from full responsibility. Most studios bring a new hire on and immediately load them with the work that justified the hire. The new person is then trying to learn the studio’s tools, conventions, clients, and processes while also delivering. The result is, predictably, slower onboarding and more first-quarter mistakes. The studio’s recent hires have spent the first three months on substantial but sub-critical work, with the full-load handover at the start of the fourth month. The shape of the first year is noticeably better.

The exit conversation is built into the start. Every hire begins with a conversation about how the relationship would end. What signals would tell either side it was time. What the notice period is. What the fair severance would look like. This sounds dark. It is, in fact, the conversation that produces the cleanest working relationship, because both sides know what the boundaries are.

The studios I respect most on this axis are the ones who hire less often than they think they should and keep people for longer than the industry average. The studios who churn through hires are usually the ones whose hiring process was not designed for what hiring actually requires.

For studios reading more on this, Basecamp’s hiring guide is the cleanest practical writing I have read on the subject. Patrick McKenzie on engineer hiring is in the same neighbourhood, more focused on software but mostly applicable. The two-week paid trial pattern is, separately, the model that 37signals has run for years.

Hiring slowly is, in the end, an admission that the cost of a wrong hire is much higher than the cost of carrying the workload yourself for an extra month while the right hire becomes available. Most small studios learn this the hard way. Worth learning it once.