Hire for agency | lau.rent

Motivation goes a long way

Fifteen years ago, I had an open position in my team. I didn’t have the budget to hire externally, and the role required rare technical skills. As a result, I didn’t have many candidates, and none with the set of skills I was looking for.

I did however have a very motivated candidate. His enthusiasm for the job, his will to join the team was so obvious that I put him in the “maybe” pile after our initial interview, despite his lack of experience in the technologies required for the job.

After that initial interview he came around to my office every few days, checking in on the hiring process, expressing his interest in the job. After a while, and with no other serious candidate in sight, I gave in to his relentless follow-ups and hired him.

And… he was great at the job. In his first few weeks, he set up a lab in his office1, self-trained on the tech stack and soon reached the level of technical fluency I was expecting for the role. His previous experiences in unrelated fields gave him a breadth of knowledge and a unique way of looking at problems that was highly beneficial to the team. He was always enthusiastic, always up for new challenges, and frequently volunteered for the less glamorous tasks.

In the 3+ years he worked for me, I never regretted my hiring decision once.

This episode changed my hiring lens. It showed me that “prerequisite” skills can be learned on the job if you’re driven enough, that a motivated individual can have outsized impact. So my heuristic became “hire for motivation”. Don’t look at the academic credentials; don’t worry too much if the candidate’s skill set is not exactly what you’re looking for. Instead, understand what drives the candidate, and whether the job and the team’s context will fuel that drive or hinder it. Gauge the amount of energy and enthusiasm they’ll bring to the job. If the person is motivated and the job lines up with their interests, the rest will take care of itself.

The “hire for motivation” heuristic served me well, especially as I moved to startups where I could rarely afford to hire for seniority and expertise. While I can’t prove the counterfactual, I genuinely think it helped me make better hiring decisions.

Oops, I did it again

Fast forward ten years and two companies. Same situation: I’m hiring for a role with a hard-to-find combination of technical skills and tool knowledge. I have a candidate who doesn’t have experience with the stack but shows enthusiasm and self-drive. I apply my “hire for motivation” rule and he gets the job.

You can guess what comes next: the candidate didn’t work out.

The motivation was real, but it wasn’t enough to get him good at the job.

He learned how to use the tools, but only in a superficial way: as we used them rather than as they could be used. He stalled out somewhere between advanced beginner and competence.

He didn’t always think through the implications of his acts, which led to a few production mishaps. And, worse, he didn’t learn from his mistakes, which led to the same mishaps reoccurring. And to frustrating discussions about not taking my feedback into account. (That’s the thing about people who don’t listen to feedback: they’ll also not take feedback about taking feedback 🤦)

That person never grew into the job; more than that, he demonstrated that he couldn’t grow into the job, no matter how much effort I put into it. I fired him, and my “hire for motivation” heuristic took a hit.

Get my agent on the line

It’s 2026, so every blog post must talk about AI in some way (bear with me, I promise this is relevant to our story).

Using Claude Code, especially with the more recent models, is a pleasure. Give it a task and some pointers, leave it alone, and wham: when you check back in, it’s done. It will keep working until it finishes the task (or you run out of credits). I’ve seen Claude Code do impressive feats of troubleshooting to get to the bottom of an issue.

Of course, you have to check its work. It gets things wrong. It will sometimes misunderstand your instructions, or outright lie about accomplishing the task; and when you confront it, it will insist you’re wrong. All of this is fine. I’ve been lied to before. I’ve seen my instructions being creatively interpreted. And when I give out a task, I damn well know to check the results. That’s called “being a manager”.

Working with Claude Code feels a lot like managing a smart, eager junior engineer2. You know the kind: full of enthusiasm, always picking up new stuff, excellent problem-solving skills. Still unpolished, a little raw around the edges; sometimes you wish they’d show less cleverness and more common sense. But! Tell them something once, and you don’t have to tell them again. They listen. They learn. They improve. I’ve managed a few people like that, and it’s always been a pleasure.

Empiricism state of mind

So what did...