Is It a Playbook?
A test to determine whether what you are writing is a playbook or something else wearing the label. Something else is fine. Mislabelling it is not, because readers arrive with playbook expectations and leave with instructions they will misapply.
The five-question test
Take any section of the document and ask:
1. Can I decompose it into plays? A play has four parts: entry conditions, an action, signals, and a when-not-to. If the content resists that decomposition, it is not a playbook. Prose that explains, argues, or persuades is an essay. Prose that lists steps without conditions is a procedure.
2. Does each play say when it applies? Not the topic it covers. The situation a reader would be standing in. "For teams adopting OKRs" is a topic. "When your quarterly goals are set but nobody can say which work serves them" is an entry condition. If the reader cannot match their situation to the play in under a minute, the situating has failed.
3. Does each play say when it does not apply? This is the sharpest single discriminator. Checklists, procedures, and methodologies almost never contain a when-not-to, because they assume their own applicability. A play knows it is contextual. No when-not-to sections anywhere in the document is close to a proof that it is not a playbook.
4. Could a reader tell whether the play is working? Every play names an observable signal. If the document's success criteria are "alignment", "buy-in", or "a culture of X", it has no signals, only aspirations. A signal is something a reader could point at within weeks of running the action.
5. Is judgement still required? A playbook assumes a capable reader choosing among plays based on what they see. If the document could be followed by someone with no judgement, it is a procedure or a runbook. That is not an insult. Runbooks are the right tool when the domain is stable and the steps are deterministic. Playbooks are for domains where the reader must read the situation first.
Scoring: all five yes, it is a playbook. Any no, it is one of the neighbours below.
The neighbours
What the document actually is when it fails the test, and where each one belongs.
| It is a... | Because... | Right tool when... |
|---|---|---|
| Checklist | Items to verify, no situating, order barely matters | The failure mode is forgetting, not misjudging |
| Runbook / procedure | Deterministic steps, judgement not required, one situation | The domain is stable and the steps are known |
| Methodology | A total system claiming broad applicability, adopt it wholesale | Rarely. Be suspicious of anything you cannot adopt one play at a time |
| Pattern language | Plays exist but the value is in their relationships and generative sequence | The reader is designing something, not responding to a situation |
| Style guide | Rules about form and consistency, not actions in situations | The problem is coherence across many authors |
| Essay / position piece | It argues a view, decomposition into plays fails entirely | You are changing minds, not equipping hands |
| Values statement | Aspirations without actions or signals | Almost never on its own. Values without plays are slogans |
The methodology row deserves its own warning. A methodology is what a playbook becomes when the when-not-to sections are deleted and the plays are made mandatory. That is a one-way door, and commercial pressure pushes through it because certainty sells better than judgement.
Fast heuristics
Rules of thumb when you do not want to run the full test.
The random page test. Open the document anywhere. Within one page you should hit an entry condition or a signal. If you hit three paragraphs of context-free advice, it is an essay.
The steal test. Could a team take exactly one play, run it without adopting anything else in the document, and get value? Playbooks are stealable one play at a time. Methodologies are not.
The disagreement test. Where does the document expect the reader to exercise judgement against it? A playbook has seams where the reader is supposed to push back. A procedure has none and should not.
The embarrassment test. Does it contain anything the author would rather have left out - a force, a cost, a condition that makes a play less impressive? Real plays have grain. If everything in it flatters the approach, it is marketing.
The retirement test. Could any part of this plausibly be marked retired later without the whole thing collapsing? Playbooks are collections and survive amputation. Methodologies depend on every part at once, which is why they never retire anything.
The verb test. Scan the headings. Playbook headings contain actions: notice, extract, propose, retire. If the headings are nouns - principles, pillars, dimensions, framework - the document is probably describing rather than equipping.
Using this on your own draft
Run the five questions on each section separately, not the document as a whole. Most real documents are mixed: two genuine plays, an introduction that is really an essay, and a values statement in a play costume. The test does not tell you to throw the non-play material out. It tells you to either convert it into plays, move it into an explicitly different section, or publish it as the essay it wants to be.
The one thing not to do is leave it mislabelled. The label sets the reader's contract, and a reader who runs your essay as a play will fail with it, and blame the right target.
Licensed under CC BY-SA 4.0.