The People Analyst Guide to Nine Lies About Work. 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; the substance is ours. Research anchors verified on read.
What the book argues
The second lie is "the best plan wins." The book's counter is that the detailed plan is obsolete the moment it meets reality, and that what actually wins is the best intelligence — accurate, current information moving fast to the people doing the work, so they can adjust in real time. The planning ritual (the annual deck, the cascaded targets set a year out) gives an illusion of control; the live read of what's actually happening is the thing that lets an organization respond.
What the research actually says
Well supported, with a boundary to respect. The support: in volatile environments, deliberate, detailed plans underperform adaptation — Mintzberg's distinction between intended and emergent strategy, and Weick's work on sensemaking (organizations cope by continuously making sense of unfolding events, not by executing a forecast) both land here. Forecasts degrade with horizon and volatility; the value of a year-out plan decays fast. The intelligence-beats-plan claim is, for most knowledge and operational work today, correct.
The boundary: this is not "don't plan." Capital allocation, hiring ramps, and anything with long lead time still need a plan — you can't nowcast your way to a data-center build. And intelligence is only an advantage if it reaches someone with the authority to act on it; real-time information flowing to a decision-maker who has to wait for the annual cycle is just faster frustration. So the honest version pairs two things: a live read of reality and decision rights pushed to where that read is. There's also a discipline the book doesn't name — knowing when more information is actually worth gathering before you decide (the value-of-information question), rather than waiting for certainty that won't come.
How you run it
Shift investment from the plan to the read. Build a continuous signal of what's actually happening (operational and people-side), get it to the people who can act, and push decision rights down to meet it. And spend on information deliberately — gather more only when the next datum would actually change the call.
The analysis you can run
A pulse / signal + decision-timing analysis — survey-orchestrator for the continuous read,
forecasting for the value-of-information layer (when is more data worth gathering before deciding;
EVPI/EVSI). Together they replace "execute the annual plan" with "maintain a live read and decide when
the read is good enough to act." (The VoI machinery is the same one the wizards use to keep estimates
honest.)
The AI-era turn
This is the lie AI most directly rewrites. Continuous, low-cost intelligence — nowcasts, live reads, anomaly flags — is exactly what AI is good at, so the plan-versus-intelligence balance tips further toward intelligence than 2015 imagined. The catch is the rater problem again: an AI nowcast arrives confident and unlabeled, so calibrate it (does it actually predict?) before you route decisions to it. Real- time and reliability-tested, not real-time instead of trustworthy.
What to do Monday
- Stop over-investing in the year-out plan for fast-moving work; invest in the live read instead.
- Stand up a continuous signal and get it to people who can act on it — push decision rights to the information.
- Before gathering more data, ask the value-of-information question: would the next datum change the decision?
- Treat any AI nowcast as a rater — calibrate it before you trust it.