The Principles
How we think about measuring people, and AI
A small set of essays that articulate the philosophy behind the work — why the portfolio exists, how we hold ourselves to account, and what we refuse to do. Read them as one argument, in order, or pick the one that speaks to your problem.
A note on the deal: each piece is written to be worth your time on its own — a usable piece of measurement method, sourced and defended — whether or not you ever touch one of our products. The self-interest is honest and light; the value should land regardless.
- 01
Thesis · the AI × people-analytics spine
The Other Direction
Everyone says AI will transform people analytics. The under-told story is the reverse — an AI judgment is a measurement, and the century-old science of trustworthy measurement is exactly what AI lacks. The empty direction is where the moat is.
Read →
- 02
Thesis · the North Star doctrine
The Number That Can Sink You
The test of a real metric isn't whether it can rise — it's whether it can fall and indict you. Every product here declares one before/after number on the customer's decision error, and lets that number grade it.
Read →
- 03
Thesis · source superiority
Opinions Are Free Now
When a confident, citation-shaped answer costs nothing to produce, "sounds right" stops being evidence. The only claim left worth anything is one you can trace to a graded source — and that depth is the moat.
Read →
- 04
Methodology · protected feedback
Safe to Say
You can't measure the truth about an organization with an instrument people don't feel safe answering honestly. Protection isn't a privacy setting on the survey — it's the precondition for the signal.
Read →
- 05
Thesis · the adaptive magazine
A Poverty of Attention
The problem was never a shortage of content — it's that more content floods the scarce resource. Simon named it in 1971. The cure isn't a bigger library; it's one that spends your attention on the right next thing.
Read →
- 06
Methodology · value of information
The Error Bar Is the Product
Analytics sells answers, but a point estimate hides the only thing a decision-maker actually needs: how much to trust it — and what tightening it would cost.
Read →
- 07
Methodology · measuring leadership
Measured Against Reality
We judge leaders by how they're perceived — 360s, reputation, room presence — not by the measurable state of the systems they're responsible for. The second is harder, slower, and the only one that should decide who leads.
Read →
- 08
Transparency · AI × people analytics
Show Your Work
The contest over AI in organizations will be won on legibility, not accuracy — and a century of procedural-justice research already settled why.
Read →
These are the "about us" pieces, deliberately kept in one place so the magazine stays about the field, not about the firm. The claims behind them are traceable in the proof graph.