Writing

My work gives me an unusual view of the engineering market. I spend a lot of time speaking with engineers, founders and hiring teams, and working on searches where the reality is often more complicated than the original job description.

Every fortnight, I review the useful observations from those conversations and from the searches I'm working on. I look for repeated behaviour, contradictions between candidates and companies, changes in how people work, and hiring decisions that don't fit the usual explanation.

I use AI to help review my notes, connect related conversations and identify possible patterns. It also helps me organise the evidence and produce an initial draft.

It does not decide what gets published.

I choose which observations are worth pursuing, check whether they rely too heavily on one person or company, look for counterexamples, decide what I think they might mean, and edit every piece before it appears here. Nothing is published automatically.

Conversations are anonymised by default. Names, companies and identifying details are removed unless somebody has explicitly agreed to be identified.

These pieces are observations from inside the market, not research papers or claims that every company is behaving in the same way. Where the evidence is mixed, I'll try to leave it mixed.