Podcast
Hiring is slow. What a good embedded pod actually costs you
A CTO and our delivery lead compare augmentation against in-house hiring, honestly.
A conversation about the arithmetic nobody puts in a proposal: the months a role sits open, the ramp time you pay for twice, and the point at which augmentation stops being the cheaper answer.
We are not neutral on this — we sell one of the two options. So the useful part of the conversation is where our delivery lead argues against himself: the cases where hiring in-house is plainly correct, and the engagement shapes we decline because they would not work.
The short version: augmentation wins on speed and on optionality, in-house wins on retained knowledge, and the crossover is somewhere around eighteen months for most teams. Anyone who gives you a more confident number is selling something.
More reading
Other things we have been wrong about.
Your modernisation isn't a rewrite. It's forty small evictions.
Why the strangler pattern keeps beating the big-bang replatform, and how to sequence the first six months.
Nobody asks the cost-per-task question until month four
A practical unit-economics model for agent workloads, and the three places spend quietly runs away.
Let's begin
Disagree with any of it? Tell us.
The best conversations we've had started as an argument about a post.