Talent Talks, formerly The Talent Revolution

The episode contains a handful of genuinely useful TA operational ideas - level-of-effort planning, sourced-candidate quality-of-hire correlation, splitting recruiters by function - but roughly half the runtime is personal icebreakers, guitar-building anecdotes, and mutual back-patting, which severely dilutes the substantive content per minute.

“there's a project management measurement system that we use quite a lot here at Scede is called level of effort, right? So you can calculate exactly what metrics you need to do, what activity, proactive activity you need to do”

“splitting recruiters per function. It's easier for recruiters to make relationships if they're consistently working in one business unit”

The observation that job descriptions are usually wrong and therefore AI screening on JDs is dangerous is the most counterintuitive point made; everything else - train your TA team, don't rush to fill roles, AI should assist not replace - is well-worn conventional wisdom in the talent space.

“most of the time job descriptions are wrong. That's why we calibrate with managers and we do kick off calls at the beginning and you find out half the stuff in the job description they don't actually want or don't need”

“use AI to make your job easier, but not to do parts of your job”

Nathan is a credible TA practitioner with genuine embedded recruiting experience across industries and a broad advisory portfolio, but he is a manager-level operator at a mid-size firm - not a CPO, CHRO, or someone who has scaled a talent function at a name-brand company - limiting the depth and authority of insight he can offer.

“How many companies like Scede have worked with over 250 different clients across the world. How many have helped scale businesses to IPO”

“I can't count the amount of wildcard kind of candidates that are outside of a remit that I've placed. The people that I think some recruiters would look at and just be like, no”

The episode offers only a handful of loose numbers (7 - 9 reqs per recruiter, '30 something' as overloaded, 250 clients, 3 - 6 month quality-of-hire reviews) and uses Apple/Klarna/Amazon as throwaway name-drops; there are no client case studies, dollar figures, conversion rates, or named outcomes that a practitioner could act on.

“for a recruiter to do everything they need to do, it is probably seven to nine roles a recruiter”

“I know some recruiters, not with clients, but friends of mine that have got 30 something headcount each”

Tom asks a decent follow-up on quality-of-hire measurement and pushes the AI topic with his email-client analogy, but the overall tone is a promotional partner chat where the host repeatedly validates, adds his own anecdotes, and never pushes back on a single vague or unsubstantiated claim.

“Do you have a standard quality of hire sort of measure or calculation that you use with all clients or do you work backwards from each individual client's definition of what that means?”

“What's real? What's noise?”