The Josh Bersin Company

The episode contains a few genuinely useful, non-obvious ideas - particularly the AI query for finding someone who worked at a target company at the same size/stage, and the point that time-to-fill metrics create perverse incentives - but these are buried in stretches of generic 'AI augments humans' and 'hire for fit not just talent' commentary that adds little.

“you don't want the person who's doing it now, you want the person who did it at your size and stage and help you grow to that point. So an AI query we can do is help me find someone who worked at this company when it was at this size in this role, what are they doing now?”

“If you just stop it when person's in the seat, you are not measuring anything but speed. And it actually creates the wrong incentives because you will hire fast, but you won't hire good.”

The stage-specific talent sourcing query is a moderately fresh framing, but the bulk of the episode recycles well-worn takes: AI as co-pilot, hire for fit not talent, start with the outcome not the tool - none of which is contrarian or first-principles thinking in 2024.

“it's very easy to say, oh, look at all the cool things this technology can do. Now let's find a use case for it. And then you build and adapt your processes to where the technology is.”

“The value proposition is not in replacing as many talent professionals. It's about amplifying their ability, right?”

Matt Hoffman is a genuine practitioner - head of talent and a named partner at a real seed/Series A VC firm actively working across a portfolio - which gives him credibility, but he is not a widely recognised operator who has scaled something exceptional himself, and the conversation doesn't draw out depth that distinguishes him from a competent talent consultant.

“I am a partner of Talented M13, a venture firm that focuses on seed and series A businesses.”

“there's four of us in total supporting close to 100 portfolio companies”

The guest names specific tools (FindEm, Mento, and a transcribed-as-'Breitheyer' tool), gives a concrete team ratio (4 people, ~100 companies), and uses the 10-person headcount-increase example, but there are no outcome metrics, retention numbers, placement rates, or dollar figures to back up the claims.

“We use a tool called FindEm.”

“We use Breitheyer for something like that and be able to really look over those screens over time”

The host asks reasonable scene-setting questions but consistently validates rather than probes, offering no pushback on vague or unsubstantiated claims and frequently affirming the guest mid-thought; there is no productive tension or follow-up that forces the guest to go deeper.

“I loved what you said here because these examples really show that this is not just doing things faster, right?”

“Totally agree.”