# Deployment Toolkit II — Diagnostic Craft (W-Series, D-Series, M-Series, I-Series)

> Part of the public canon of *Perceptive Innovation* by **Caleb Kow** —
> canonical URL: https://perceptiveinnovation.com/canon/toolkit-diagnostic-craft.md · index: https://perceptiveinnovation.com/llms.txt
> © 2026 Caleb Kow, all rights reserved. Quotation with attribution to Caleb Kow / perceptiveinnovation.com is welcome.
> Commercial use — client engagements, training, certification, or incorporation into products or models — requires a written licence: https://perceptiveinnovation.com/terms

Source: *The Perceptive Innovation Deployment Toolkit* (Second Edition, 2026) by Caleb Kow. This file carries the audit-grade diagnostic depth: how scores are earned, read, contested, prescribed, measured, and joined to the classic strategy frameworks. Use it to sharpen every scoring conversation, not only formal audits.

## W-06 — The Evidence Ladder (Use Everywhere)

Classify every piece of evidence by strength, strongest at the top: **R1 observed behaviour** (what adopters do: usage, retention, repeat purchase, workarounds) · **R2 unprompted verbatims** (what they say when nobody asked: tickets, reviews, forums, sales-call notes) · **R3 solicited responses** (what they say when asked: surveys, interviews, NPS commentary) · **R4 proxy metrics** (what correlates: traffic, sign-ups, press sentiment, analyst tone) · **R5 internal belief** (what the team assumes: founder conviction, engineering pride, sunk cost). **The rule:** a sphere scored only on rungs 4–5 is marked low-confidence whatever the number says — high Trust resting on claims nobody can independently verify is the engineered shape of the book's cautionary case; treat it as fragile until harder rungs confirm it. Evidence beats altitude: a rung-1 observation from the most junior person outranks a rung-5 conviction from the most senior. Tag the rung of every score basis in every output.

## W-01 / W-02 — Surfacing the Reflex (Prepare Stage)

**W-01 Technology-First Reflex Test:** ten statements rated 1–5, completed individually and anonymously, only the aggregate revealed (e.g. "when adoption lags, our first instinct is to add or improve features"; "wins are attributed to the product; losses are attributed to the market"; "'if we build it well, adoption will follow' would pass unchallenged in our meetings"). Bands: 10–24 perception-aware · 25–39 technology-leaning · 40–50 technology-first. Facilitator's line: "Nobody in this room chose this bias. It is what competent technical organisations are trained into. The next hour is about seeing it, not confessing it." **W-02 Adoption Post-Mortem:** one past disappointment; the technical case for why it should have won; what actually happened in numbers; the harvest — every reason intended adopters did not embrace it, in their words; which reasons the team already knew before launch; what it cost in money, time, credibility. Cluster the harvest into at most six piles, unnamed — the facilitator reveals the spheres only after the clusters are on the wall; the room discovers its own story already sorts into them. Recognition, not instruction.

## W-03 / W-04 — Framing and Scoring

**W-03 Audit Framing Sheet** locks the unit of analysis: the innovation in one sentence, no adjectives · **one primary audience — name the person, not the segment** · audit occasion (pre-launch gate / live triage / retrospective) · the named decision this audit informs, with date and owner · sponsor sign-off · secondary audiences parked, each earning its own sheet later · sealed priors (each scorer's private composite guess before evidence opens). No named decision, no audit. **W-04 PIPAS Scoring Sheet** anchors: **1 = this sphere works actively against you · 3 = neutral or uncertain · 5 = this sphere works actively for you.** Per sphere: the score, the **strongest single piece of evidence**, and confidence **firm/thin**. The six question forms (use these everywhere): Identity — *what is this, and do I become someone I want to be by adopting it?* · Value — *do the benefits outweigh the costs: money, time, learning, switching?* · Trust — *will it perform as promised, and will my decision be vindicated?* · Risk — *what could go wrong, and how reversible is it if it does?* · Effort — *how much work to understand, adopt, and keep using it?* · Future — *is this the direction things are heading, or a dead end?* Composite bands: below 15 severe · 15–19 significant · 20–24 moderate · 25–30 likely adoption.

## D-01…D-06 — Sphere Deep-Dives (Anchors · Probes · Patterns · Affinities)

Reach for the deep-dive when scoring stalls, when a sphere scores lowest, or when dissent ≥2 lands there. Score the audience's perception, not your intention.

**D-01 Identity.** Anchors: 1 audience actively disidentifies — "not for people like us" arises unprompted; adoption carries social cost · 2 identity friction — users adopt quietly, hide usage, apologise for it internally · 3 neutral — the product says nothing about the adopter · 4 positive fit — adopters volunteer it suits "teams like ours"; some self-badge · 5 adoption itself confers desired identity — users badge, evangelise, recruit. Probes: who do they picture as the typical user, and is that flattering? · what does adopting say about them to peers, bosses, customers? · which segment adopted that you did not design for, and why? · is anyone hiding their usage — who, from whom? · what identity did the category assign you before you arrived? Patterns: **the camera-on-the-face problem** (a technically defensible feature makes the adopter look wrong to their tribe — bystander perception vetoed the wearer's enthusiasm) · **borrowed identity backfire** (positioning against a loved incumbent inherits its enemies as well as its category). Lever affinities: Talent (visible, credible people whose adoption signals "people like you belong here") · Temperature (cool the mockery climate before asking anyone to badge publicly) · Thought (reframe the category so adopting reads as foresight, not eccentricity).

**D-02 Value.** Anchors: 1 costs visibly exceed benefits in the audience's own accounting — "can't justify" is the default verdict · 2 benefit real but illegible — it takes your spreadsheet, not theirs, to see it · 3 contested — champions can argue it, sceptics can argue back, neither wins in a meeting · 4 benefits legible and accepted; total cost seen as fair · 5 payoff self-evident and quotable — adopters repeat the one-sentence case unprompted. Probes: can the economic owner state the payoff in one sentence, without you? · what does it replace, and what does the replacement really cost them? · where does the value land versus who pays — same desk or different? · what would they lose if it vanished tomorrow, their words? · at +20% price who stays; at −20% who suddenly appears? Patterns: **the complexity trap** (features added to raise value raise Effort faster — perceived value falls as the spec sheet grows) · **value–payer split** (users get the benefit, a different function pays the cost — score per audience or the matrix lies). Affinities: Thought (give the market the frame in which the value computes) · Tangibility (convert claims into a demonstration the sceptic can run alone).

**D-03 Trust.** Anchors: 1 claims disbelieved by default — the maker's word actively discounts the message · 2 belief only with heavy proof burdens — every claim relitigated in every meeting · 3 willing to be convinced — trust follows whoever shows up with evidence first · 4 claims accepted; the company expected to stand behind them · 5 reputation does the selling — third parties defend the claims for you. Probes: which specific claim is believed least, and who has actually tested it? · whose endorsement would settle it, and why haven't they given it? · what happened the last time this category over-promised them? · what proof exists an adopter can verify without you in the room? · would your decision be vindicated if it appeared in the trade press? Patterns: **trust deficit inheritance** (category scandals tax every entrant — your Trace is partly the category's Trace, whether you earned it or not) · **engineered trust** (a high Trust score resting on unverifiable claims — mark it low-confidence and treat it as structurally fragile). Affinities: Trace (build a record you own: published outcomes, named customers, kept promises) · Talent (credible intermediaries — the veterinarian handing over the bottle) · Tangibility (proof an adopter can check independently).

**D-04 Risk.** Anchors: 1 failure feels catastrophic and irreversible — career-risk language appears in notes · 2 serious exposure with unclear recovery — "what if" questions dominate late-stage calls · 3 manageable with safeguards the audience must ask for one by one · 4 understood, bounded, mostly reversible; standard assurances suffice · 5 visibly reversible and safeguarded — trying it is perceived as the safe option. Probes: what is the worst credible outcome in their telling, not yours? · who personally carries the blame if it fails, and do they know it? · how reversible does exiting look — contract, data, workflow, reputation? · which safeguards exist that the audience doesn't know about? · what did the last failed vendor in this seat cost them? Patterns: **the legacy prison** (enterprise audiences over-weight switching risk because they are still paying for the last migration) · **asymmetric blame** (approving a failure is punished; rejecting a success is invisible — expect gatekeepers to score Risk harsher than users; a two in a gatekeeper cell can veto a four everywhere else). Affinities: Trace (a visible record of failures handled well beats a claim of no failures) · Trouble (meet the objections in the open — pre-answer the redlines) · Tenacity (signal you will still be here when the audit committee asks).

**D-05 Effort.** Anchors: 1 adoption demands re-learning work itself; abandonment precedes first value · 2 heavy lift — dedicated staff, long onboarding, visible disruption to current flow · 3 real learning curve, survivable with support the adopter must request · 4 low friction — first value inside days; fits existing habits with minor adjustment · 5 effectively effortless — adoption inside an existing habit; trying costs nothing. Probes: how long to first genuine value — measured, not promised? · what must the adopter unlearn, and who on their side does the work? · where in the funnel do they stall, and what were they attempting? · what does maintaining it cost after the honeymoon? · could a sceptic succeed unassisted on day one? Patterns: **the zero-effort exception** (the book's breakout case won by making trial cost nothing — a text box and a question; when Effort hits five, weaker spheres get time to heal) · **easier beats better** (the Quiet Winner shape — chronically underestimated by rooms trained to look for spectacle). Affinities: Technology (spend the technical budget on removing steps, not adding features) · Time (meet the audience at the moment the old way visibly breaks) · Training (documentation, templates and community that answer before the ticket is raised).

**D-06 Future.** Anchors: 1 perceived as a dead end — betting on it reads as career malpractice · 2 doubt dominates — acquisition rumours, thin ecosystem, roadmap questions in every call · 3 open question — the audience is waiting for a signal either way · 4 seen as on-path — direction credible, ecosystem forming, commitments lengthening · 5 perceived as inevitable — not adopting is what requires justification. Probes: where does the audience say the category is going, and are you on that line? · what signal would convince them you'll exist, improving, in three years? · who is building on you, and would they say so publicly? · what happened to the last "inevitable" thing they bet on? · if your biggest competitor copied you tomorrow, what survives? Patterns: **format-war shadow** (audiences that watched a standards war punish ambiguity — adoption stalls until the Future question resolves, regardless of technical merit) · **the slow collapse** (Future erodes quietly for years before the composite falls — which is why W-14 tracks trajectory and hooks watch this sphere between audits). Affinities: Trace (published roadmap kept, quarter after quarter — direction as a habit) · Time (tie your story to a shift the audience already believes in) · Tenacity (visible sustained investment when it would be easier to hedge).

## W-05 — Sphere × Audience Matrix

Thirty-six readings on one page: six spheres × six audience columns (**User · Market · Regulator · Internal · Competitor · Investor**), evidence noted per cell. Find **the lowest cell on the page** and ask: can this audience veto adoption? If yes, that cell outranks every other finding. Weight by leverage, not size — the audience that can block adoption outranks the audience that is merely large. The User column is the default headline — until a gatekeeper column shows a two.

## W-11 — Mental Model Canvas

Maps what the audience already believes the innovation is, because scores are read through that model, not around it. Current model: "It's basically a ______ for ______" · the category shelf they file it on · what that shelf makes them expect. Source signals that built the model: first exposure (demo? press? a rival's pitch?) · category history and its casualties (Trace) · price point and packaging signals · who they saw using it first (the identity carrier) · your own past products' shadows. Desired model: one sentence, audience language, no adjectives you'd have to defend. Then the **three first-experience moments** that will construct it — for each: where in the journey · what they must conclude · the signal we will engineer. Doctrine: you cannot install a mental model; you can only construct the conditions under which the audience builds the one you intend — mostly in the first minutes of exposure. If current and desired models sit on different shelves, expect the gap to surface as Identity or Effort on W-04 before anywhere else.

## W-12 — Perception Gap Map

Score W-04 twice for the same audience — once from internal belief only (no evidence pack), once from the evidence pack only — and plot both; the bar between the dots is the gap. Log per sphere: belief · evidence · gap · direction (optimistic/pessimistic) · first-conversation owner. The audit's most common first finding is not a low score — it is a wide gap. Optimistic gaps (belief above evidence) hide veto risk; pessimistic gaps hide under-priced strengths. The widest gap gets the first conversation; the pattern of gaps is the reflex test made visible. Gaps ≥2 open the calibration conversation; persistent gaps become C-05 assumptions.

## W-07 — The Profile Reader (Four Shapes)

Match the converged scores to a shape to read the failure mechanism at a glance. **The Builder's Profile** — Value and Future high, Identity and Effort low: the team sees the destination; the audience cannot see themselves in it (in the book: Google Glass, early Apple Vision Pro). Levers: Thought + Temperature for Identity; Technology + Training for Effort. **The Hype Profile** — Identity and Future high, Trust and Risk low: desire outrunning belief — honest when admitted, engineered when denied (early Web3; the Theranos caution). Levers: Trace, Tangibility, Trouble — build evidence you own; meet criticism in the open. **The Trust Cliff** — five spheres serviceable, Trust at two or below, usually inherited: the category's accumulated trace is toxic (Web3 after 2022; Microsoft's decade rebuild). Levers: Trace, Talent as credible intermediaries, Tangibility — demonstrations an adopter can check alone. **The Quiet Winner** — Effort at five, nothing below three, nothing spectacular: easier-beats-better, most often underestimated by rooms trained to look for drama (Slack 2014, WhatsApp, Stripe). Levers: protect it — Tenacity and a public Trace; the shape erodes quietly when incumbents respond. No clean match? Most real audits are hybrids — read the lowest sphere first, then the widest dissent. **And the all-threes profile is not a shape — it is a red flag that the room scored for harmony instead of from evidence. Send it back.**

## W-13 — Dissent Log & Monitoring Hooks

Dissent is retained information, not noise — the sphere the room cannot agree on is often the sphere that later moves. Convert every ≥2-point spread into instrumentation. Hook fields: sphere · score range · holdout (who and their score) · **the holdout's evidence, verbatim** · **observable trigger — "if we see…", measurable, dated** · owner · check date · fired? A good hook is falsifiable and boring: "if two more enterprise security reviews stall at legal by 30 September" beats "if trust gets worse." When a hook fires, the holdout was early, not difficult — say so in the room, and re-score the sphere before the quarter ends.

## W-14 — The Composite Trajectory

Plot the composite and that audit's lowest sphere across audits; log what moved since last and the lever credited or blamed. A twenty-two that has climbed from nineteen and a twenty-two that has fallen from twenty-five are different situations demanding different moves — the falling one gets triage, the climbing one gets patience. **Never report a composite without its arrow** (climbing / falling / first reading).

## W-08 — The Sphere → Lever Selector

Affinities, not assignments — evidence overrides the map. Strongest affinities (from the D-series): Identity → Talent, Temperature, Thought · Value → Thought, Tangibility · Trust → Trace, Talent, Tangibility · Risk → Trace, Trouble, Tenacity · Effort → Technology, Time, Training · Future → Trace, Time, Tenacity. (These nest inside the Field Kit's fuller Figure 11.12 mapping; use the affinity set for audit-grade selection, the full matrix for programme design.) Category emphasis in the book: consumer T1·T10·T6 · enterprise T2·T4·T5 · Web3 T2·T8 · AI T3·T7·T8. Selection ceiling: **two target spheres maximum, one lever each**, each written as a single-sentence move.

## W-15a/b — 10 T's Current State

Before any lever is pulled, assess today's practice per T: what we do today on this lever · working (yes/partly/no) · neglected entirely · maturity 1–5 · **evidence this is the state (not opinion)** · **the one gap to close first**. The kit's loudest alarm: maturity 1–2 on a lever with high affinity to your weakest sphere — that pairing is next quarter's W-09, almost by definition.

## W-09 — The Intervention Brief (The Contract)

Converts the reading into a contract: **two spheres, one lever each, one quarter.** Per brief: target sphere · score · audience · lever (one T) · the verbatim evidence that set the score · **the move — one sentence; if it needs two, it is two moves** · named owner — a person, not a function · review date within 90 days · perception metric, baseline today → target at review · **kill-or-adapt criterion — what reading at review kills or reshapes this move** · sponsor sign-off and date. Two spheres is the ceiling, not the target — one done properly beats two done politely. Nothing leaves the room without names and dates on this page.

## W-16 — The Intervention Roadmap (Four Quarters)

Sequencing rules: **Trust floors first** — amplifying Identity or Value claims above a Trust ≤2 pours water into a cracked jug; floor Trust, then amplify · **Effort before scale** — fix time-to-first-value before any push that multiplies first-time users, or you scale the disappointment · **Trace compounds** — Trace moves slowest and pays longest; if it appears in any quarter, start it in Q1 regardless · **one inherits, one starts** — each new quarter carries at most one continuing pair and at most one new. Tracks: focus pairs A and B (sphere → lever → owner) · re-audit dates · metric checkpoints · hooks under watch. The ceiling, illustrated: four quarters × two pairs = eight moves a year maximum, and most organisations should run six. The roadmap's job is refusal — write refused items in the margin; they are next year's shortlist, not this year's overflow.

## W-10 — The 90-Day Tracker

Three phases, twelve deliverables, thirteen weekly pulses; the sponsor initials each phase gate. Phase One (days 1–30) Diagnostic: stakeholder inventory across all six audiences · PIPAS applied per audience, W-05 rolled up · 10 T's current state documented · priorities set, framework signed off. Phase Two (31–60) Foundation: trust investment sprint shipped (limitations published) · identity positioning reviewed across every channel · first-experience journey mapped and instrumented · early wins captured, learning system running. Phase Three (61–90) Execution: operating cadence booked into real calendars · cross-functional perception review standing · capability documented so it survives personnel change · transformation review done, re-audit scheduled. The commonest failure is the diagnostic–execution gap: a reading everyone praises and nobody owns. If a deliverable has no initials by its gate date, escalate — do not extend.

## M-01 — Perception Metrics Selector

The value chain: perception moves → behaviour moves → the commercial number moves; leading metrics live at the first arrow, lagging at the second. Pick **one leading and one lagging per W-09 target sphere**, with baseline and target, before the intervention starts — a metric chosen after the intervention begins is a story, not a measure. Menu: **Identity** leading — share of new adopters from the intended segment, self-badging count; lagging — referral rate, community growth. **Value** leading — one-sentence payoff repeated back in calls (tracked share), discount depth to close; lagging — net revenue retention, win rate vs value objections. **Trust** leading — reference calls granted, claim-belief survey delta; lagging — pilot→production conversion, security review pass rate. **Risk** leading — safeguard-page views, reversibility asks answered pre-emptively; lagging — redline count per contract, insurance/indemnity demands. **Effort** leading — time-to-first-value, activation step completion; lagging — onboarding ticket volume, 90-day retention. **Future** leading — roadmap-question share of QBRs (falling), partners announcing on you; lagging — multi-year commitment share, analyst trajectory rating. Committed metrics table: sphere · metric (one leading, one lagging) · baseline today · target at review · source/owner. Today's baseline is next quarter's comparison — written once, used twice.

## M-02 — Perception ROI & Maturity

The ROI ledger — attribute honestly, claim conservatively: win rate on the audited segment · sales cycle length · discount depth to close · net revenue retention · onboarding cost per account · pipeline from referral/reference — each before/after/Δ/annualised value, against investment (people + programmes + tools; audit and facilitation days), with return claimed at **attribution share ≤50%**. The maturity ladder: 1 **Reactive** — perception discussed only after launches disappoint; no shared language · 2 **Aware** — spheres named; audits ad hoc; findings praised, rarely owned · 3 **Instrumented** — baselines dated; metrics chosen before moves; cadence booked and kept · 4 **Managed** — perception review inside standing processes; W-09s ship and get verdicts · 5 **Strategic** — perception evidence shapes portfolio and budget; re-audits are routine, not events. Circle where you are; box where the year ends.

## M-03 — Quarterly Re-Audit Comparison

Same audience, same anchors, fresh evidence — or the comparison is fiction. Per sphere: last · this · Δ · **why — evidence, not narrative** · hook fired? Then verdicts on last quarter's briefs: **worked / partial / kill**, each with its consequence for Q+1 (continue / replace / stop). If the audience itself changed, that is a new baseline, not a delta — start a fresh trajectory line and say so in the readout. The first re-audit is where the practice becomes real.

## I-01 — PIPAS × Disruptive Innovation

Analogy: disruption theory finds where the castle wall is thinnest — overlooked segments the incumbent is structurally unmotivated to defend. Perception decides whether the gate opens when you arrive. **Thin wall, shut gate: still no entry.** Classify the play with evidence: sustaining (perception battle: Trust and Trace against the incumbent's — expect the audience to default to the devil it knows) · low-end foothold (Value must read as smart economy, not cheap risk — Identity and Trust guard the door) · new-market foothold (the audience has no shelf for you — Mental Model and Effort decide everything). Capture the job to be done in the foothold audience's words. Score **two audiences, two readings** — foothold and mainstream; the disruption is visible in the difference. The asymmetry check: write the incumbent's perception of the foothold in their own likely words ("not our customer", "margin-dilutive", "toy") — their Identity sphere is your Time lever; the window stays open exactly as long as that sentence stays true upstairs. Overshoot, translated: "performance oversupply" is a perception event — the audience stops perceiving marginal capability as Value and starts perceiving it as Effort and cost.

## I-02 — PIPAS × Blue Ocean Strategy

Analogy: the strategy canvas charts unclaimed water; the audit checks whether your intended passengers believe that water is navigable — a blue ocean the buyer cannot perceive is a red line on the P&L. Every ERRC action lands on spheres — write the perception consequence: **Eliminate** → Effort ↓ and Value clarity ↑, but check Identity — removed factors may be badges the audience wears · **Reduce** → Value recomputes; verify with survey evidence that buyers, not engineers, priced the factor · **Raise** → Trust must carry the claim; a raised factor unbelieved is Temperature, not Value · **Create** → no shelf exists; created factors enter through the Mental Model Canvas — name the model before you price the factor. Non-customer tiers: for each (soon-to-defect · refusing · unexplored), name the **blocking sphere** holding the tier out. Value innovation is the strategy; *perceived* value innovation is the adoption — run W-04 on the drawn curve with each target tier before committing the cost structure; the veto is cheapest to hear on paper.

## I-03 — PIPAS × Crossing the Chasm

The product does not change mid-leap; the audience reading it does — the chasm is a perception break, and scoring both sides makes it a number. Visionaries buy the future and forgive the present; pragmatists buy references and whole products. Score each sphere for both audiences and log the gap, with the typical flips: Identity — "bold" becomes "unproven for firms like ours" · Value — future payoff → present, referenceable payoff · Trust — your claims → their peers' testimony · Risk — tolerated → career exposure; reversibility rules · Effort — pilot heroics → must fit existing workflow · Future — vision → evidence of staying power. Whole-product check: integrations the pragmatist assumes · implementation and support partners · standards/compliance fit · training and documentation depth · references in their niche. Beachhead selector: choose the pragmatist niche with the **smallest Trust gap and the densest peer-reference network**, then win it whole before widening — bowling pins fall sideways through references, not forward through features.

## I-04 — Choosing the Lens (The Combined Workflow)

What each theory answers and leaves open: *Where can we attack; whom first?* → Disruptive Innovation — gives attack vectors, footholds, incumbent asymmetries; leaves open whether the foothold's perception admits you. *What offer shape creates demand others ignore?* → Blue Ocean — gives value curves, ERRC, non-customer tiers; leaves open whether buyers perceive the new curve as value. *Why did early wins stall before the mainstream?* → Crossing the Chasm — gives adoption stages, whole product, beachhead logic; leaves open the size of the gap, sphere by sphere. *Will this audience let any of it in?* → PIPAS + 10 T's, always — gives the reading, the veto risk, the levers; leaves open the strategic move itself — bring a lens above. **The combined workflow: STRATEGY LENS (Christensen · Kim & Mauborgne · Moore choose the move) → PIPAS GATE (W-04 per target audience scores whether the move can land) → 10 T's (W-08/09 shift the odds where the gate flagged risk) → CADENCE (A-01/M-03 re-read the audience as the move unfolds).** The classics choose the move. The reading scores the odds of the audience admitting it. The levers shift those odds. None of the three lenses is displaced by the fourth — a perception audit run without a strategy lens produces a well-measured drift, and a strategy executed without the reading walks confidently into a veto.