THROUGHLINE — The Lenses (operational levers)

The four research disciplines as operational levers — checklists and diagnostic moves, not literature. Conforms to THROUGHLINE.md. Status: DRAFT v0.1. Core rule (THROUGHLINE §3): lenses describe a venture’s properties; they never grade it on desirability. A “red flag” here is always incoherence or self-deception, never an unflattering-but-owned property. Every lever ends by forcing ownership, not improvement.

A. IO-PSYCH lens — motivation, identity, people

Hardest at Layers 0 (Substrate) and 3 (Identity).

Levers

Motivation audit (Self-Determination: autonomy · competence · relatedness; intrinsic vs extrinsic). Name the actual driver. Then: “What would make you walk away?” — the walk-away set is the real constraint set.
Identity-claim test. Does the stated identity match revealed behaviour and the incentives the ARTIFICE creates? (A “creator-first” identity with a creator-hostile cap table is incoherent.)
Role & selection. Who must exist for this to run; what’s the founder-fit; what’s deliberately not staffed.
Culture-as-incentive. What behaviour does the structure actually reward (vs claim to)?

Red flags (defects): stated motivation ≠ real (infidelity); identity claims the ARTIFICE contradicts. Not a defect: an unusual, selfish, or non-altruistic motivation — if owned.

B. BEHAVIORALISM lens — choice architecture, loops, ethics

Hardest at Layers 6 (Mechanism) and 4 (Brand).

Levers

Funnel friction map. Where is friction along the engagement path; is “free”/“easy” actually frictionless, or just labelled so?
Default & choice architecture. What is the default path? Opt-in vs opt-out? What does the architecture make easy (that is what will happen)?
Habit loop (Fogg B = MAP: Behaviour needs Motivation + Ability + Prompt; or cue→routine→reward). For each desired behaviour, confirm all three exist. A mechanism that wants a behaviour with no prompt or no ability is assuming, not designing.
Incentive / mechanism design. Is it incentive-compatible? Where will it be gamed? (e.g. governance-token whale capture.)
Ethics-of-nudge line. Surface any manipulation. Does engineering compulsion or a dark pattern contradict the stated IDENTITY? If yes, that’s a coherence defect, full stop.

Red flags (defects): a mechanism assuming behaviour with no M/A/P and the gamble unowned; dark patterns that contradict declared values. Not a defect: low conversion, a small/niche funnel, a deliberately high-friction “earned” experience — if owned.

C. ECONOMIC-RESEARCH lens — structure, defensibility, unit economics

Hardest at Layers 1 (Thesis) and 5 (Artifice). This is the lens most prone to smuggling in desirability — it is strictly diagnostic here.

Levers

Value-pool map. Where does money/attention actually accrue, and is that where the THESIS says it does? (Descriptive — mismatch is a coherence flag, not a quality one.)
Defensibility characterization. Name the moat — network effects · IP · switching costs · brand · regulatory · none. If “none,” confirm the founder accepts the competitive exposure as a chosen property. “No moat” is a fact to own, not a failure to fix.
Viability vs. maximization (the desirability trap). Two distinct reads: (a) viability — can it sustain itself on the founder’s own success terms? A genuine “no” (losing money per unit while intending to profit) is a real finding. (b) maximization — is it leaving margin / scale / exit on the table? That is not a defect. Arizona Iced Tea’s 99¢ is the model: superb viability, deliberately un-maximized. Surface the gap to “optimal-by-desirability”; never tell the founder to close it.
Desirability-pressure exposure. Does the chosen ARTIFICE (Layer 5) import an external success definition that can override the founder’s? Public markets and VC carry a growth/margin-maximization mandate — the force that gutted profitable game studios for not being maximally desirable. Name the exposure; insulation (private · steward-ownership · mission-lock) is legitimate and often the point.
Commons / public-goods read. If the core is given away, name what is actually scarce and monetizable, and whether that split is deliberate. (OPEN PANEL: free story, owned universe.)
Mechanism design. For any market/marketplace/token: is the rule-set incentive- compatible and sybil/gaming-resistant?

Red flags (defects): claiming a moat that isn’t there; assuming profitability you can’t trace; conflating revenue with defensibility; a value-pool that contradicts the thesis. Not a defect: declining to maximize — no moat by choice, un-maximized margin (Arizona’s 99¢), no exit, deliberately small/niche, insulating from public-market pressure. A real finding (not a desirability call): genuine un-viability on the founder’s own stated terms.

D. IDENTITY / SEMIOTICS lens — archetype, narrative, naming

Hardest at Layers 3 (Identity) and 4 (Brand).

Levers

Archetype pick. Choose the character (the 12 brand archetypes, or a bespoke one). Every downstream expression checks against it.
Narrative spine. Protagonist, tension, stakes. Can the venture be told as a story in three sentences?
Naming & namespace clearance. Run every name through tools/landgrab (domains, ENS, socials, tickers, packages) — Layer-4 CLEAR move.
Distinctiveness / trademark exposure. Generic ↔ arbitrary spectrum; descriptive marks are weak. (LONGBOX is generic-in-comics — a described property, fine for a sub-brand.)

Red flags (defects): a name/visual that contradicts the archetype; an uncleared collision shipped anyway; identity↔brand mismatch. Not a defect: a weird, polarizing, or deliberately non-mass-market brand — if it expresses the identity faithfully.

Using the lenses

Run them in the GROUND move (THROUGHLINE §5) after generating a layer. The output of a lens pass is a short list of named properties + owned/not-owned — never a score. Anything “not-owned” goes back to the founder as a forced choice (interview), then to the source-of- truth as a decision. Properties the founder owns are done, however unattractive.

DRAFT v0.1. Conforms to THROUGHLINE.md. Diagnose, don’t approve.