Four questions [Adler frame]
Q1. What is it about? Operational and strategic principles derived from two decades inside Facebook’s growth organisation — from user zero to over a billion users. Gleit presents a set of durable frameworks for product thinking, team operations, and people management under conditions of rapid scale.
Q2. How is it argued? Through the “Naomi-isms” — a named, transferable set of principles including canonical thinking, the PM-as-conductor model, growth accounting, and a distinctive people typology. Each ism is grounded in observed failures at Facebook and articulated as a corrective.
Q3. Is it true? The canonical thinking principle is well-evidenced: version proliferation is one of the most documented failure modes in fast-moving teams. The PM-as-conductor analogy is illuminating but has limits — a conductor does not own outcomes the way a PM does. Growth accounting is a sound diagnostic tool. The disagreeableness typology is observationally sharp.
Q4. What of it? The isms are immediately portable. Canonical thinking alone would improve most teams. Growth accounting should be standard practice for any consumer product. The people typology reframes how to think about “supportiveness” in management.
Glossary
Canonical thinking. The discipline of maintaining a single authoritative version of every document, decision record, and communication thread. Eliminates version-reconciliation as a cost of operations.
Growth accounting. The formula: net active users = new users − stale users + resurrected users. Disaggregates net growth into its components, making retention collapse visible even when new-user numbers look healthy.
PM as conductor. The model of a PM as responsible for the performance of the whole, without playing any individual instrument. Authority derives from accountability for outcome, not from functional expertise.
Disagreeableness-giver. A person who provides honest, sometimes unwelcome feedback in service of the recipient’s growth. Contrasted with the agreeable giver who sounds supportive but withholds truth.
Canonical thread. A single designated communication channel for a topic, to which all relevant parties are directed. Eliminates parallel threads and the reconciliation cost they impose.
Canonical thinking
The core operational principle: for every topic — document, decision, meeting, chat thread — maintain one canonical version. [§ Canonical thinking]
The failure mode it corrects: version proliferation, where multiple slightly-different documents or threads coexist, and team energy goes into reconciling them rather than using them. At scale, this failure mode is severe — hours per week per person spent resolving which version is current.
Implementation: designate the canonical artefact explicitly and enforce the norm that updates happen there and only there. Any time someone creates a parallel version, redirect immediately.
The principle extends to meetings: one canonical meeting per topic, with a single owner and a standing agenda, rather than ad-hoc scheduling that produces inconsistent decision records.
Growth accounting
Standard growth metrics obscure the underlying health of a user base. [§ Growth accounting]
The formula disaggregates net growth:
- New users: first-time activations
- Stale users: previously active users who have churned
- Resurrected users: churned users who have returned
A product with flat or growing MAU can be in structural trouble if stale users are accumulating faster than new ones arrive, and resurrected users are masking the underlying retention collapse. Growth accounting makes that structural problem visible.
Gleit credits this framework with enabling Facebook’s growth team to diagnose and address retention issues that would have been invisible under standard MAU reporting.
PM as conductor
The PM does not play an instrument — they are responsible for the performance. [§ PM as conductor]
This framing clarifies a common confusion: the PM’s authority is not grounded in being the best engineer, the best designer, or the best analyst. It is grounded in being accountable for the product’s outcome. That accountability requires the ability to hold the whole in view — to hear what the ensemble is producing and make adjustments — without collapsing into any single instrument’s perspective.
Implication: the PM who argues from their own technical or design preferences has confused their role. Their job is to make the ensemble produce the right result, not to express their own preferences through the output.
The people typology
Gleit’s typology on the giver-taker axis: [§ People typology]
- Disagreeable giver: tells you the truth regardless of how it lands. Highest value. Looks difficult to manage but makes direct reports better.
- Agreeable giver: sounds supportive, validates frequently, but withholds important negative feedback. Feels good in the short run; fails their reports in the long run.
- Disagreeable taker: demands and criticises but does not invest in others.
- Agreeable taker: pleasant but self-serving.
The insight is that “supportiveness” as commonly understood conflates the agreeable giver with the disagreeable giver. The agreeable giver’s apparent support is a form of neglect — it deprives the direct report of the feedback they need to improve.
Problem space → opportunity → execute
Gleit’s three-stage product thinking sequence: [§ Problem to execution]
- Understand the problem space thoroughly — what is actually happening and for whom.
- Identify the specific opportunity — the narrowing from problem space to addressable gap.
- Execute.
Most teams skip to stage 3. The result is technically competent work on the wrong problem, which is worse than slow work on the right problem because it accumulates technical and organisational debt.