Will Larson: evaluation

Transcript context

…So then going back to your example of the hiring pipeline, let's come back to that to help connect this definition, which I've never heard, which is awesome, very clear to how you actually implemented say in hiring. The first thing to do is to get a model out there of any sort. And so you think about your stocks. And so in a hiring pipeline you might have potential candidates and this could be basically infinite, and then you have a couple of inflows. So you could have sourced, you could have outreach, you could have referrals, and then you have the candidates and those... How many people get sourced. There's probably a function of number of sources you have dedicated to a role. So you'd have another stock of how many sources you have and then that would impact the rate going from potential candidates to actual candidates you've sourced. And then from candidates that you have in that box, you'd have a conversion rate for people who pass the first recruiter screen and that would move into step two of your process. And then from step two you have maybe a hiring manager screen and that would be another conversion rate. So you can see over time how the candidates of the infinite potential candidates like wandering around LinkedIn, posting about their deep thoughts, convert into actual people, your first, your second, your third. But then as you get deeper into it, you start to actually see interesting things. And so for a lot of candidate or a lot of pipelines, the biggest issue is hiring managers don't want to extend any offers because the hiring managers can't get to confidence on any candidate. And you'll see in this pipeline, you'll see a ton of candidates getting to offer stage, but almost none of them converting from a potential offer to actually offer. And then you can say, "Hey, here's the problem. You need to go work with the recruiter and the hiring managers on getting conviction about who they should hire. Classic problem with early managers, right? Here's a second problem. Manager wants to extend a ton of offers. They do extend them and none of them actually accept. And so that focuses you on the second problem. But there's a third potential world where actually you're just not getting enough candidates in. You're actually doing a great job of making decisions, great job of closing candidates. There's just not enough candidates coming in. And so by looking at this and you can build this model, then you can go to your applicant tracking system, a greenhouse or whatever and pull the historicals and you can just see how the historicals work versus how you'd expect it to work. You can see the drop-offs and this helps you figure out where should you go try to fix things first. I think we've all worked in companies where you roll out big changes with no data behind them because like, "Oh, it feels like we're not hard enough in how we evaluate candidates or something. " You go change a bunch of stuff. But often the real problem might be that the hiring managers making offer extensions to people who never passed the loop anyway. It's just the managers are issuing too many offers because they're panicking. ten the real problem might be that the hiring managers making offer extensions to people who never passed the loop anyway. It's just the managers are issuing too many offers because they're panicking. And man, less true now, but a decade... Or not a decade, two years ago, hiring managers panicking to get offers out, that was a real thing that happened a lot. And this just helps you take a complex kind of abstract problem and turn it into something you can actually work in a systematic way.…

Stored transcript either side of the excerpt. The highlighted words are the published quote; the surrounding text is unedited source, never generated.