METHODOLOGY // REVENUE RISK

Every number we publish survives a CFO.

The methodology is published. The inputs are transparent. The assumptions are named. The customer can override any variable. Transparency is the moat. Vendors hide how they use your data. Blackout shows how we calculate its cost.

01 // PHILOSOPHY

Every number must survive a CFO's scrutiny.

We work to four principles. Each one is a discount on the number, not a multiplier. Conservative by design.

We quantify what's quantifiable.

Per-vendor revenue at risk, calculable from contract value, data classes, and observable behavior.

We estimate what's estimable.

Campaign effectiveness degradation. Renewal asymmetry. Bands with confidence intervals.

We reveal what can only be shown.

OAuth access to internal email. The finding is the number. No estimate needed.

We never inflate.

Every variable has a discount built in. Caps everywhere. Ranges instead of point estimates.

02 // FOUR TRACKS

Revenue risk decomposes into four independent tracks.

Each maps to a Collapse Engine subsystem, a different audience, a different level of quantifiability. Each compounds independently.
Track 01Calculable

Competitive Signal Leakage

SubsystemTHE BROKER
ChannelCAC Subsidization
AudienceCFO, CEO

We're funding our competitors

Track 02Estimable

Campaign Effectiveness Degradation

SubsystemTHE ORACLE
ChannelSignal Corruption
AudienceCMO, VP Marketing

That's why CAC keeps climbing

Track 03Binary

Corporate Intelligence Exposure

SubsystemTHE REAPER
ChannelGTM Attack Surface
AudienceCISO, General Counsel

They can read WHAT?

Track 04Calculable ceilings + Binary

Regulatory & Legal Tail Risk

SubsystemTHE COUNSELOR
ChannelLegal Tail Risk
AudienceGC, CFO, DPO

Our DPA says X. Runtime shows Y.

▸ The Collapse Engine describes WHY revenue degrades. The methodology calculates HOW MUCH.

Subsystem
Revenue channel
Track
THE ORACLE — Collapse of Truth
Signal Corruption
Track 2
THE BROKER — Collapse of Control
CAC Subsidization
Track 1
THE REAPER — Collapse of Safety
GTM Attack Surface
Track 3
THE COUNSELOR — Collapse of Legitimacy
Legal Tail Risk
Track 4
03 // TRACK 1 · THE EQUATION

Three variables. One per-vendor result.

Each variable is independently observable or estimable. Customer-overridable. Conservative by design.

▸ Core equation · per vendor

Revenue at Risk = DAV × CEF × EC

DAV

Data Access Value

What the data this vendor can access is actually worth, anchored to the customer's own contract value.

Contract value × data contribution ratio. Or, fallback: records × published reference rates.

CEF

Competitive Exposure Factor

What percentage of this vendor's data access creates competitive advantage for your rivals.

Shared model (binary) × competitor overlap × relevance multiplier. Capped at 60%.

EC

Exfiltration Confidence

How confident we are that this vendor is actually extracting and using this data versus merely having theoretical access.

Confirmed (90–100%) → Probable (60–80%) → Possible (20–40%) → Not Observed (0%).

▸ Worked example

6sense on a $40M ARR B2B SaaS company

50K monthly visitors · Salesforce integration · $85K/yr contract

▸ Step 1 · DAV (Data Access Value)

Browser-side

Visitor identity

50K/month × $0.25/record

$150,000
Authenticated

CRM contacts

8,200 × $1.50/record/month

$147,600
Authenticated

Pipeline / deals

340 active × $5.00/record/month

$20,400
Total DAV$318,000

▸ Step 2 · CEF (Competitive Exposure Factor)

Shared intent graph: YES

Competitor overlap: 4 of 8 known competitors = 50%

Relevance multiplier:

Raw: 150% → capped at 60%

Applied CEF× 0.60

▸ Step 3 · EC (Exfiltration Confidence)

Browser identity: Confirmed (95%)

CRM pull: Probable (70%)

Blended (DAV-weighted): 80%

Applied EC× 0.80

▸ Result

$318,000 × 0.60 × 0.80 = $152,640 / year

$130K – $175K

estimated annual revenue at risk from 6sense

Range reflects EC confidence band. Customer overrides will tighten the estimate further.

04 // TRACKS 2, 3, 4

The other three tracks. Different math, same rigor.

Track 02 · Campaign Effectiveness Degradation

Your forecast is built on signal you can't trust.

Estimable

Vendor scripts manufacture identity events your analytics ingests as real. The same vendors take credit for the pipeline they helped fabricate. Channel mix, attribution model, and CAC calculation read off corrupted data.

▸ At-risk campaign spend

Campaign Spend (vendor-flagged accounts) × Signal Compromise Ratio

If 30% of a vendor's signal derives from pooled customer data, 30% of campaign spend targeting that vendor's surfaced accounts is built on compromised signal.

Track 03 · Corporate Intelligence Exposure

They can read your CEO's emails.

Binary

Vendor access to executive correspondence, deal strategy notes, competitive positioning, and pricing decisions through CRM integrations approved for “product functionality.” Not calculable. Either they have access or they don't.

▸ The Product Functionality Test

For every OAuth scope a vendor holds: what specific, documented product feature requires this data class?

Delta between scopes requested and scopes justified by documented features = the unexplained access surface. The customer draws the conclusion.

Track 04 · Regulatory & Legal Tail Risk

Your DPA says X. Runtime shows Y.

Observation only

The measurable gap between governance artifacts (DPAs, privacy policies, consent mechanisms, subprocessor disclosures) and what actually happens at runtime. Surface name in product: Regulatory Touchpoints.

▸ Posture

We observe behavior and cite the regulations or statutes that address that behavior pattern. We do not assert violations, applicability, or liability.

Blackout provides the evidence. Counsel provides the assessment.

05 // PRECISION LAYERS

The number tightens as you connect more.

Each layer adds inputs. Each layer narrows the band. The estimate becomes a measurement at Layer 4.

Layer 01

Scan-only estimate

Wide range

Vendor count, VIDB behavioral profiles, public competitive overlap

$180K–$340K annual exposure

Layer 02

OAuth-enriched

Tighter range

+ actual OAuth scopes, CRM object counts, API pull frequency

$130K–$175K from 6sense

Layer 03

Contract-enriched

Narrow bands

+ contract values, stated terms, subprocessor disclosures

$42K subsidization on $85K spend

Layer 04

Enforcement-verified

Measured, not estimated

+ actual exfiltration events observed and blocked by the diode

4,218 events blocked. $95K verified savings.

▸ Customer override principle · every input is editable · numbers become collaborative

06 // CONSERVATIVE BY DESIGN

Every cap is a discount on the number.

Five mechanisms that keep estimates honest. Disclosed in every calculation. Visible in every output.

Every variable carries a discount

EC confidence levels: 0% – 100%, never above observed evidence

CEF capped at 60%

No competitive-exposure factor exceeds 0.60, regardless of inputs

Time multipliers capped at 1.5×

Integration age longer than 24 months never multiplies above 1.5×

Cross-vendor correlation capped at 1.3×

Aggregate stack risk never multiplies above 1.3×

Ranges replace point estimates

Wherever confidence is below Confirmed, output is a band

07 // WHAT WE DON'T CLAIM

The methodology is honest about its limits.

We do not claim vendors are "selling" specific customer data. We show what they can access, what their products require, and the gap.

We do not claim precise dollar losses. We show estimated exposure ranges with published methodology.

We do not claim attribution fraud without direct evidence. We identify the conditions that make circular attribution possible.

We do not assign intent. A vendor with unexplained OAuth scopes may have legitimate reasons we haven't identified.

We do not provide legal advice. Regulatory exposure ceilings use published penalty structures applied to observed evidence.

We do not claim DPA discrepancies constitute violations. We show the delta. The legal determination is your counsel's.

See your number. Then verify it.

Run a scan. Get a Layer 1 estimate in 60 seconds. Connect your CRM and contracts to tighten the band. Override anything you disagree with.

▸ Methodology v0.2 · 2026-04-22 · Conservative by design