The People Analyst Guide to First, Break All the Rules. Format: what the book argues → what the research actually says → how you run it → the analysis you can run → the AI-era turn → what to do Monday. No reproduction of the book's text — and never the Q12 items (proprietary). The substance is ours; research anchors verified on read.
What the book argues
The book opens by arguing that you can't manage what you can't measure, and that the right thing to measure isn't satisfaction or perks — it's a small set of workplace conditions that predict whether a team will produce: clarity about expectations, having what you need, a chance to do what you do best, recognition, someone who cares, room to grow. Get those conditions right locally and the outcomes follow. The provocation is that this short, mundane-sounding list outperforms the elaborate climate surveys it replaced — because it measures the conditions a manager can actually change.
What the research actually says
This is one of the better-evidenced claims in the book. The engagement-to-performance link has been tested at scale: Gallup's own meta-analyses (Harter and colleagues) aggregate across thousands of business units and find that units scoring higher on these engagement conditions tend to score higher on productivity, retention, and customer outcomes. Independent of the proprietary instrument, the broader literature agrees on the shape — work engagement (the Job Demands–Resources tradition: Bakker, Schaufeli, Demerouti) predicts performance and retention, and the conditions that drive it are local and manageable. (Read-verify the specific effect sizes before quoting them; correlations are real but modest, and the causal direction runs both ways.)
Two honest notes a people analyst has to add. First, the Q12 itself is Gallup's copyrighted instrument — this Guide measures the constructs (the conditions), with our own items; we don't reproduce theirs, and neither should you. Second, the measuring stick is only useful at the team level (the Nine Lies Lie 1 point) and only if you trust the measurement (the Lie 6 point) — a company-wide average of a noisy survey is the false precision both of those chapters warn about. The measuring stick is right; most organizations swing it at the wrong altitude.
How you run it
Measure the conditions, with your own items, at the team level, often enough to act. Define the handful of conditions that a manager can move (expectations, resources, strengths-use, recognition, growth), measure them where they vary (the team), and check the reliability before you report (Lie 6). The point isn't a score; it's a per-team read a manager can do something about.
The analysis you can run
A team-level engagement analysis — survey-orchestrator for the condition measures (ours, not the
Q12), segmentation-studio to resolve teams and surface the within-company spread. It returns each team's
condition profile and the gap to outcomes, so leadership acts on the specific teams that need it rather
than a company-wide program. (Same machinery as Nine Lies Lie 1; reliability-checked per Lie 6.)
The AI-era turn
The annual, company-wide engagement survey is the artifact this chapter accidentally spawned at most companies. AI makes the better version cheap: a continuous, team-level read of the conditions, light enough to run often, with the noise handled. The discipline is the rater discipline — an AI-summarized engagement read is still a rater; calibrate it before you route decisions to it.
What to do Monday
- Measure a handful of manager-movable conditions, in your own words — not the Q12 (it's proprietary).
- Report at the team level; lead with the spread across teams, not a company average.
- Reliability-check the read before anyone acts on it.
- Point each manager at their team's condition gaps — that's the whole value of the measuring stick.