DiagnoseLink 1 — Demand + Source

Lead Source Quality Audit

Determine whether each lead source is producing economically valuable opportunities, not merely cheap leads. The Audit evaluates every source through the full downstream Revenue Chain so you can tell a source problem…

Link

Link Name

Demand + Source

Playbook

Demand + Source Playbook

Loop Step

Diagnose

Role

audit

Purpose

Governing Principle

A source can look successful at the cost-per-lead (CPL) level while quietly destroying value downstream. Do not judge source quality by CPL alone. Judge it by the economics of the opportunities the source produces through the entire Revenue Chain.

Evaluation Chain

Lead Volume
Qualification
Booking
Show
Close
Revenue
Gross Profit

How To Use

1. Pick a measurement window long enough to accumulate outcomes (leads take time to become retained clients). Use the same window for every source so comparisons are fair. Record the exact window on each row.
2. Create one row per source at the grain you actually control (typically source + channel + campaign). Do not blend two campaigns into one row.
3. Fill the raw input columns from your ad platforms, call tracking, forms, and CRM. Where a number is unknown, enter 'Unknown' — never guess and never enter 0 to fill a gap.
4. Compute the derived metrics with the formulas below. Any derived metric whose denominator is Unknown is itself Unknown.
5. Set attribution confidence and data-quality confidence for the row, and apply the sample-size rule to every rate.
6. Read the row through the evaluation chain (Volume -> Qualification -> Booking -> Show -> Close -> Revenue -> Gross Profit), then classify the source using the patterns below.
7. Write the source-specific leak observation: the single earliest point in the chain where this source's economics break down.
8. Hand the completed Audit to the Lead Source Scorecard & Budget Decision Matrix (RLR-08) to convert findings into a Scale / Maintain / Repair / Reduce / Stop decision.

Columns

Column	What it means
Source name	The traffic source as you would name it to a colleague (e.g. Google LSA, Meta Prospecting, Referral Partner A).
Channel	The medium class (paid search, paid social, LSA/marketplace, organic, referral, offline, direct).
Campaign	The specific campaign or line item inside the channel.
Measurement window	Exact date range the row covers. Identical across all sources being compared.
Spend	Fully-loaded cost attributable to this source for the window (media + platform fees + agency/portion where applicable). Referral/organic may be 0 or a partner fee.
Inquiries / leads	All raw inquiries the source produced (calls, forms, chats, messages), before any quality filter.
Valid inquiries	Inquiries that are real potential buyers (excludes spam, wrong-number, robocalls, duplicates, existing clients, job seekers).
Contacted leads	Valid inquiries your team actually reached (two-way contact), not merely attempted.
Qualified leads	Contacted leads that meet your qualification standard (fit + intent + eligibility). Use the same qualification definition across sources.
Booked appointments	Qualified leads that scheduled a consultation/appointment.
Shows	Booked appointments that actually attended.
Retained / closed clients	Shows that became paying/retained clients.
Revenue	Booked/collected revenue attributable to retained clients from this source in the window. State whether it is booked or collected and keep it consistent.
Gross profit (where available)	Revenue net of direct delivery/fulfillment cost. Enter Unknown if you cannot compute it reliably rather than approximating.
CPL	Cheapness of raw inquiries. The most over-trusted and most misleading single metric. Low CPL means nothing if the leads do not convert.
Cost per qualified lead	First quality-aware cost. A source with low CPL but high cost-per-qualified is producing volume without fit.
Cost per booked appointment	Cost to generate a real, scheduled opportunity.
Cost per show	Cost to get a qualified buyer in the room. Sensitive to no-show behavior, which can be a booking/reminder problem rather than a source problem.
CAC / cost per retained client	Cost to acquire one paying client. The primary economic verdict on the source, provided attribution is trustworthy.
Lead-to-qualified rate	Raw quality of what the source sends. Low here points to a source/targeting problem.
Lead-to-appointment rate	How efficiently inquiries become opportunities. Low with healthy lead-to-qualified points downstream (speed-to-lead, booking friction).
Show rate	Confirmation/reminder health and buyer commitment. Usually a Booking + Handoff (Link 5) issue, not a source issue.
Close rate	Sales-conversation strength for buyers from this source. Low with good upstream quality points to a sales/offer problem, not a source problem.
Revenue per lead	Top-line economic value of a raw inquiry from this source. Directly comparable across sources.
Gross profit per lead	The truest single measure of source value when available: what one inquiry is worth after delivery cost.
Attribution confidence	How sure you are the outcomes on this row really came from this source.
Data-quality confidence	How clean and complete the underlying counts are (dedup, spam removal, consistent definitions).
Sample-size warning	Flag set when a rate is built on too few events to be conclusive.
Source-specific leak observation	One sentence naming the earliest point in the evaluation chain where this source's economics break down, and whether it looks like a source problem or a downstream conversion problem.

Source Classification

High-volume / low-quality

Signals: Low CPL; high leads; low lead-to-qualified rate; high cost-per-qualified
Likely Meaning: The source sends many inquiries that do not fit. This is a source/targeting problem.
Recommended Next Step: Tighten targeting/offer/keywords or reduce; verify it is not a qualification-definition problem before cutting.

Expensive but economically strong

Signals: High CPL; strong lead-to-qualified and close rate; low CAC relative to revenue/gross profit per lead
Likely Meaning: The source looks costly per lead but produces profitable clients. Cutting it on CPL alone would destroy value.
Recommended Next Step: Consider MAINTAIN or SCALE in RLR-08; protect from CPL-based cuts.

Good leads, poor downstream conversion

Signals: Healthy lead-to-qualified rate; weak booked/show/close rates; low retained despite good qualified volume
Likely Meaning: The source is fine; the leak is downstream (Speed-to-Lead, Follow-Up, Booking + Handoff, or sales).
Recommended Next Step: Route to REPAIR of the downstream link — not REDUCE/STOP of the source. Confirm with the relevant Link 2-5 audit.

Broken attribution

Signals: Leads present but revenue/retained largely Unknown; attribution confidence Low; large 'unknown source' bucket elsewhere
Likely Meaning: You cannot yet judge this source because outcomes are not linked back to it.
Recommended Next Step: Fix tracking with RLR-09 first; mark decision INSUFFICIENT EVIDENCE until attribution confidence improves.

Insufficient data for a confident decision

Signals: Sample-size warning set on the deciding rates; very low denominators; new source with short history
Likely Meaning: Direction may be visible but the numbers are not yet conclusive.
Recommended Next Step: Keep measuring to a defined minimum sample; decision is INSUFFICIENT EVIDENCE, not STOP.

Source Vs Conversion Problem

Principle: The Audit's most important job is separating a SOURCE PROBLEM from a CONVERSION SYSTEM PROBLEM.
Read Earliest Breakpoint: Walk the chain in order. The FIRST rate that collapses tells you where the problem lives.
Rules: If lead-to-qualified is weak, the problem is likely the SOURCE (wrong people).; If lead-to-qualified is healthy but lead-to-appointment collapses, suspect Speed-to-Lead (Link 2) or booking friction (Link 5), not the source.; If show rate collapses, suspect confirmation/reminder and handoff (Link 5), not the source.; If close rate collapses for good-fit, shown buyers, suspect the sales conversation/offer, not the source.; A source that delivers good-fit leads into a broken downstream system should be marked REPAIR (downstream), never punished as a bad source.

Missing Data Rules

Unknown is a valid, honest entry and is never the same as 0.
A derived metric with an Unknown denominator is Unknown; do not infer it.
A row with Unknown revenue/gross profit cannot carry an economic verdict (CAC/GP-per-lead) — it can only carry upstream quality signals plus a measurement action.
Never backfill a gap with an assumed or industry number to complete a row.

Sample Size Guidance

Policy: Rates built on very few events are directional, not conclusive. Set a minimum denominator appropriate to your volume before you treat a rate as decisive.
Convention: A common statistical convention is to treat any rate whose denominator is under ~30 as directional only. This is a data-confidence convention you set for your business, NOT a ShiFt-approved performance benchmark.
Action: When the deciding rate is below your minimum, set the sample-size warning and route the decision to INSUFFICIENT EVIDENCE in RLR-08 rather than SCALE/REDUCE/STOP.

Outputs

A per-source verdict located on the evaluation chain (where value is created or lost).
A source-vs-conversion-problem classification for each source.
A confidence-qualified read (attribution + data-quality + sample size) on every conclusion.
A ranked list of sources to hand to RLR-08 for a budget decision, plus a list of rows blocked on measurement.

Guardrails

No fabricated benchmark thresholds. Where no approved threshold exists, compare sources to each other and to their own history, and state the basis.
No invented client results or revenue claims.
CPL is never sufficient grounds for a verdict on its own.
Every economic conclusion must trace to counts with a stated data source and confidence.

Feeds Into

Resource Id: RLR-08
Title: Lead Source Scorecard & Budget Decision Matrix™

Remeasure

After any source or downstream repair, re-run this Audit over a comparable window before reallocating substantial budget. Installation of a tracking or targeting change is not proof the source economics changed.

How to implement

Work through every row honestly; enter "Unknown" rather than guessing.

How to verify

A diagnosis is only useful if its inputs are trustworthy — note attribution and data-quality confidence.

Rather not do this part yourself?

You can repair your Revenue Chain yourself with this Kit — or bring in ShiFt at any point. You do not need to finish the DIY path first.

Done with you

Revenue Leak Scan

A ShiFt specialist runs the diagnosis with you and hands you a prioritized repair plan.

Get My Revenue Leak Scan

Done for you

ShiFt RevenueOS

ShiFt installs and operates the full Revenue Chain system for you — ConvertOS included.

See RevenueOS