Cut / Keep / Fix / Scale Decision Matrix

Id

RLR-37

Link

6

Link Name

Attribution + Decision Intelligence

Role

control

Loop Step

Prioritize

Purpose

Turn measurement into an explicit, accountable management decision about each source and process. The Matrix evaluates evidence across the whole Revenue Chain — not one top-of-funnel metric — and returns a named decision with an owner, a horizon, and a review date. It reuses the approved prioritization method (RLR-06) and introduces no new opaque ranking formula.

Governing Principle

Do not cut a source or process because one metric looks weak. A campaign that produces strong-fit prospects the business then responds to slowly and books poorly is not a source problem — it is a Revenue Chain execution problem, and the decision is FIX CONVERSION SYSTEM, not STOP SOURCE. Every decision distinguishes a SOURCE PROBLEM from a REVENUE CHAIN EXECUTION PROBLEM, and INSUFFICIENT EVIDENCE is always a valid decision.

Reuses Priority Method

RLR-06

Reuses Priority Method Note

Ranking of which decision to act on first reuses the Revenue Leak Priority Matrix (RLR-06) — financial impact, evidence confidence, operational severity, dependencies, effort, risk. This Matrix decides the disposition per source/process; RLR-06 sequences the resulting actions. No competing scoring formula is introduced.

Canonical Actions

• SCALE
• KEEP
• FIX
• REDUCE
• STOP
• INSUFFICIENT_EVIDENCE

Decision States

Evidence Considered

• lead_volume
• lead_quality
• response_performance
• qualification_performance
• follow_up_performance
• booking_performance
• show_performance
• close_rate
• cac
• revenue
• gross_profit_where_available
• attribution_confidence
• sample_size
• operational_leakage
• current_repair_activity

Source Vs Execution

Distinction

A SOURCE PROBLEM means the demand itself is wrong (poor fit, uneconomic even with sound execution). A REVENUE CHAIN EXECUTION PROBLEM means the demand is adequate/good-fit but a downstream link loses it. The Matrix requires evidence across the chain before attributing a weak result to the source.

Example

A campaign produces strong-fit prospects, but the business responds slowly (Link 2) and books poorly (Link 5). The default decision is NOT STOP SOURCE — it is FIX CONVERSION SYSTEM (route to the Speed-to-Lead and Booking Playbooks), then re-evaluate the source once execution is verified.

Guardrails

• Never reach STOP on cost-per-lead, low volume, one weak top-of-funnel metric, or a small sample alone.
• INSUFFICIENT_EVIDENCE is a valid, expected decision; weak or incomplete data must not be forced into a management action.
• A source is not cut for a downstream execution leak; fix the execution first (FIX), then re-evaluate the source.
• Every decision records evidence, owner, horizon, review date, success criteria, and the next decision — no opaque calls.

Measured Vs Modeled

Rule

Decisions rest on MEASURED evidence. Modeled figures (Revenue Leak Calculator) may size the opportunity but never stand alone as the evidence for STOP/REDUCE/SCALE; they are labeled modeled and paired with measured attribution confidence.

Closed Loop

Note

Decisions are the hinge of the closed loop. A SCALE/REDUCE/STOP decision feeds source allocation back to Link 1 (Demand + Source); a FIX decision feeds process repair, staffing/ownership, automation, workflow, follow-up, booking, or measurement priorities back to the specific leaking link. Link 6 informs Link 1 and every leaking link — the chain does not end at the decision.

Compliance Guardrail

Decisions use only data the business is permitted to hold and act on under its privacy and consent policy, and never rely on bypassing provider restrictions or opt-outs. Attribution limits are treated as evidence constraints, not ignored. Jurisdiction- and industry-specific requirements are applied by the operator.

What To Measure

For each source and process: fit-adjusted economics across the whole chain (volume, quality, response, qualification, follow-up, booking, show, close, CAC, revenue, gross profit where available), attribution confidence, and sample size — enough to assign SCALE/KEEP/FIX/REDUCE/STOP/INSUFFICIENT_EVIDENCE.

Data Required

Reconciled chain-level metrics from the Dashboard (RLR-36), defined by the KPI Dictionary (RLR-35), with the attribution-confidence level from the Audit (RLR-33).

How To Interpret

Weigh evidence across the chain and separate source problems from execution problems before deciding. A weak downstream metric with good-fit demand routes to FIX, not STOP. Low confidence routes to INSUFFICIENT_EVIDENCE, not a guess.

Action That Follows

Sequence the resulting actions with the Priority Matrix (RLR-06), assign owners and repairs on the Repair Plan (RLR-39), and feed source/process decisions back to Link 1 and the leaking link (closed loop).

Owner

A named commercial/RevOps decision owner runs the Matrix; each decision has an accountable action owner.

How To Implement

For each source/process, populate the evidence fields from the dashboard, classify source-vs-execution, select the decision state, and record evidence/owner/horizon/review-date/success-criteria/next-decision. Review in the Weekly Operating Review (RLR-38).

How To Test

Take a real recent source that looked weak on cost-per-lead and run it through the Matrix; confirm the logic routes a good-fit-but-poorly-converted source to FIX rather than STOP, and that a low-confidence case returns INSUFFICIENT_EVIDENCE.

When To Remeasure

Re-evaluate each decision at its review date and at the 14/30-day re-diagnostic (RLR-42), confirming the decision's success criteria were met before locking it in.

How To Know It Improved

The Matrix is working when decisions are evidence-backed and documented, sources are no longer cut for fixable execution leaks, and reallocated investment produces better reconciled revenue — not because a decision was made quickly.