00:14:36 · Siadhal Magos, Sebastian Johnson
We take you through a flow to share the right context with the agent such that it essentially knows how to communicate in a way that is representative how you like to communicate. And, really, that's what whenever you're doing a hiring, like, you're a founder, if you're a hiring manager, if you're a recruiter, if you're out there trying to get the right button seats, you're thinking about who are the right people and how do I wanna speak to them. And if I could times that by ten and have all of my people do it for me, then I would, and that's sort of what this is an answer answer to. [00:15:06] S0: And what have been some of the early results? Because you've been running for more, I guess, like, in a sort of beta mode or with some some existing customers for a little bit before you've the public launch. What's been the reaction for people who have been using it?
00:15:18 · Siadhal Magos
[00:15:18] S1: It's been phenomenal, frankly. And, like, the real proof becomes when you start to see... A lot of times recruiting functions, will look at their recruiting data, you know, periodically, let's say, once a quarter, and they look at... Oh, and also, like, where what are our source... What are the... What's our source of hire? Oh, it turns out we did a lot of sourcing or we did a lot of inbound, I e, people came to us this quarter. And the real proof in the pudding is gonna be when you... Is is when you start to see the number of hires that are attributed to fillmore, like, become significant, and it becomes significant quite quickly. There's usually about a six to eight week delay after adopting the product to starting to see actual hires made as a result just because there's still interview process and general sort of courting of the candidate you wanna do. That's a very human thing. But when you start to see change, that's obviously a game changer. And that's and that's when people start to ask the more transformational question of, like, is this the Claude code moment for recruiting? Because there's, like, clearly highly economically valuable work that got done that we didn't do and is now affecting how much we're winning at recruiting. So so that's that's like the big payoff. What you see though is as soon as people start adopting the product and they start to get meetings in their calendar, you know, despite not having to sort of toil away themselves, that you see that delight. So early results are we we close about 15 to 20% of the roles that are added to the platform in the first eight weeks when we onboard. Well, currently, we're on track to onboard about 350, maybe 400 roles per month, and this is it from quite a small team. So, like, when you think about the sort of the the multiple you're getting on human effort, essentially, it's a small team small team of agent operators on our side. Fillmore is working four, 500 new roles every month. You know, you're talking about, like, that's sort of a scale of a very large agency at that... At this point. Yeah. And what we... The the the best sign of all is obviously customers that add more roles. Right? You might start with, hey. I'll try that...
00:17:24 · Siadhal Magos, Sebastian Johnson
Try this out on a couple of a couple of roles to get started, and then you start to see these, you know, world class companies like Replit or Luma AI or ModelML, and they decide to upload 10, 20, 30 roles because they realize this is... This should just be a part of how they make every hire. [00:17:41] S0: And so, Siadhal, you mentioned a bunch of metrics in there. When you are looking at the health of the recruitment processes of the club, you know, the 7,000 companies that you're you're working with, what are some of the the most important metrics that shows you that that things are going well, that either their recruitment process is going well or they're using their products well? Those are really, like, this is the the KPI, the north star that we that we're looking at.
00:18:03 · Siadhal Magos
[00:18:03] S1: Yeah. Well, we focus mainly on, like... There's two questions there. Like, what do we... How do we know that, essentially, we're being helpful to our customers? We're helping our customers win. And the second is, how would a customer understand it relative to their peers if they're actually doing... If they're winning at recruiting? Those are those are related but slightly different. Yeah. How do we understand... What what do we look for? We primarily look at usage. The reality is you could... We could we could deploy Metaview to a to a, you know, a tech company, and they could still hire without using it. Right? There... The hiring has existed before Metaview existed. There's a way to do it without Metaview. It's it's not it's not sort of... Yeah. The the the the evidence that we're helping you achieve your goals, we think the truest form is that people are essentially opting into going to this thing to do the work instead of other systems. So we primarily look at usage, and we look at that across, are you using multiple of our agents or just one? And we... Multiple agents is better because it means you're making better use of the context. And so that's what we we sort of... That's what we know makes customers successful. It's it's slightly different measure versus what... How a customer is measuring their own success at recruiting as a whole and what what almost usage of Metaview should ladder to, which essentially comes down to the things that we we care about on their behalf there are time to fill. So how quickly do we go from creating a role saying, hey. We have this need that we need to solve by bringing talent into the company to having someone at least with an accepted offer and pref... Preferably even their their butt is in that seat. Time to fill, we think, is really what many of our customers already do, but we think that's the thing that should change most as a result of of AI. And mileage varies on this. I don't wanna act like Metaview sort of does it for you. It's... This is still very much as workflow where the humans involved is super important, but we see averages around 25, 30% reduction at the moment in time to fill. We think there's a lot more headroom there. We think we could bring that down. We we think it should become essentially, you know, as unrecognizable as an engineer's job is where they can, you know, close...
00:20:10 · Siadhal Magos, Sebastian Johnson
You know, complete about eight to 10 x as much work, we should be seeing similar. You know, recruiters should be running 20 roles at a time instead of five or seeing, you know, multiples more as many candidates as they are per role. There's some variance where they should be able to spend more attention. Yeah. Their human attention in order to get either better or faster output. [00:20:30] S0: And how how does Metaview compare on some of those metrics to the clients that you work with? Whether it's, time to get bummed seat.
00:20:38 · Siadhal Magos
[00:20:38] S1: Yeah. We are... Our customers are very high performing companies to be to be clear on on average, but we're we're we're tracking at about, I think, a twenty three day time to fill, which is very solid. That's like... That's low, basically. Again, varies across different job values. Most companies will will aim... Be aiming for... They'll see good as around sort of thirty to forty five days time to fill. So we're we're... Especially with the impact of fillmore over the last three months or so, we've we've been using fillmore internally for about three It's... That's massively brought it down. I mean, I'm literally getting off... Once we get off this, I've got a call with a candidate that I wanna make an offer to that was a fillmore candidate. I... Fillmore find... Found them and engaged with them, and I was sort of going through our... Almost our communication history, which, you know, I was aware of because, obviously, that's why film will work, keeps you aware of how it's going. But, you know, it's... The the reality is there's this conversation that had occurred with this candidate that primed them to get ready to meet with me that was not, like, from my fingers to their ears, you know, or from my fingers to their eyes. Measure. So it's real gay. It just means you... You're sort of like your surface area is much like... Anyway, to answer your question, we're top quartile. Time to hire isn't necessarily the type of thing. You could have, like, a one day time to fill if you wanted to, if you were willing to, you know, hire the first person you met, but that's obviously not. So so you wanna be careful. You wanna countermeasure to these... Counterweights to these measures, but we're we're doing well.