Funnel operating rules

Current working standard: 16 October 2026. Local documentation update; not published.

Use this for new funnels. The Karada plans and framework lessons are reference examples, not universal prices, section orders, conversion benchmarks or launch settings.

Start here

1. Define the buyer and the transaction

Before designing, write one answer per field:

An enquiry, qualification call, event registration and purchase are different transactions. A free fit call must not imply a free workshop, booked appointment or completed task analysis.

2. Design the whole journey

Give each stage one owner, a completion condition and a recovery route. The page is only part of the funnel.

3. Give each page section one job

• Answer a buyer question; do not print an introduction announcing that question.
• Let logos, an example or a deliverable carry the message when they are clearer than prose.
• Keep captions that establish ownership, source, scope or consent.
• Place relevant evidence near the claim it supports.
• Repeat the same primary action where useful; do not introduce competing promises.
• Keep necessary policy, accessibility and source links usable.
• Adapt the sequence to the offer and visitor knowledge. A particular 9- or 15-section layout is not universal.
• Use PAS/AIDA and story frameworks to organize a hypothesis, not to justify invented pain, statistics or urgency.

4. Use an offer ladder only where it helps

Acquisition.com’s official course distinguishes attraction, upsell, downsell and continuity offers. Our operational interpretation:

• A qualification call is not automatically a paid entry product.
• A newsletter requires its own subscription choice; an enquiry is not automatic marketing enrolment.
• Never make an optional upgrade appear necessary to finish an already accepted registration.
• Allow a genuine decline. Do not copy guilt-based decline labels.
• Do not add subscriptions to a one-day service without a distinct continuing benefit.
• Prices, bundles and scarcity from someone else’s launch are not our defaults.

5. Keep proof honest

No fabricated recent-purchase feeds, unsupported rankings, invented reference prices, resettable deadlines or unapproved guarantees. A source’s claim is not independent verification of that claim.

6. Specify the paths after the form

Reminder timing depends on the meeting lead time and chosen channel. The teardown’s example cadence is not a default automation. A resource, reminder sequence or response-time promise needs an accountable owner before it appears in customer copy.

7. Separate receipt, fit, attendance and payment

• Maintain qualification, appointment and payment state separately. Attended does not mean qualified; qualified does not mean paid.
• A retry preserves the original submission identity and packet. Count one accepted application, not every attempt.
• A repeat customer can have multiple legitimate applications or orders. Do not collapse all history solely by phone number.
• Carry stable offer, funnel, variant and source identifiers through the operational handoff.
• Keep visitor-entered values out of general analytics events. Follow the site’s existing allowlist and consent contract.
• Browser and server reports of the same conversion must use the configured deduplication contract.
• Send later quality feedback only from a trusted CRM decision. Do not label unqualified enquiries as qualified conversions.
• Do not turn a booking, browser return URL or submitted form into a paid conversion.
• For Stripe, follow the official fulfillment guidance; verify payment and process fulfillment idempotently.

No CRM, ad, consent or payment setting is changed merely by adopting this guide.

8. Measure outcomes with explicit denominators

Use consistent cohorts, maturity windows, attribution rules and currency. Include no-shows, duplicate events and refunds in the relevant reconciliation. Front-end sales revenue alone does not establish acquisition profitability; delivery costs and refunds matter.

Unknown measurement is marked unknown, never green or zero. Qualitative page readiness is not a conversion-performance score.

9. Run interpretable experiments

1. Write the hypothesis, primary metric, guardrails and decision rule before starting.
2. Choose the unit of assignment and keep it stable.
3. A headline test changes the headline while preserving offer, traffic policy and form behavior.
4. A new buyer, offer or journey is a whole-funnel experiment. Do not attribute its result to the headline alone.
5. Different ad sets pointing to different URLs can test commercial packages, but do not by themselves establish randomized landing-page causality.
6. Use the approved platform experiment or controlled assignment for a causal comparison. Do not silently change an active campaign to introduce it.
7. Derive sample size from the baseline, useful detectable effect and chosen uncertainty. There is no universal 100-conversion minimum or guaranteed two-week answer.
8. Include qualification lag and relevant calendar variation. Declare inconclusive results when the evidence is insufficient.
9. Keep a stable version record and do not edit an active variant mid-test.
10. Keep new offers and payment flows in review until their delivery and legal/payment gates are satisfied.

10. Review in this order

1. Confirm buyer, promise and actual transaction.
2. Map the accepted, declined, no-response, no-show and unqualified paths.
3. Review the complete mobile reading sequence and each section’s job.
4. Check proof attribution and approval of commercial claims.
5. Verify validation, draft retention, retry identity and accessibility.
6. Trace one authorized test record through CRM and measurement.
7. Verify payment and delivery separately if payments are enabled.
8. Establish the baseline before testing more variants or increasing spend.

Record implementation, local verification, publication, downstream delivery and owner acceptance as separate states.

References and boundaries

• Public FigJam reconstruction: observed read-only in the browser on 16 October 2026. It is a third-party map, not an audited revenue record.
• Official Money Models course categories: primary source for the offer-type vocabulary.
• Russell Brunson’s book descriptions: audience, message and funnel frameworks; not a source for our conversion forecasts.
• Karadau funnel explanation: segmentation and follow-up reference. Its displayed performance claims were not verified by this review.
• Existing live playbook hub: current browser entry point. Local edits in this guide are not claimed to be published there.

The owner’s pasted teardown transcript is summarized in the Devtx evaluation. Its launch-revenue estimates and speculative premium upsell are excluded from our evidence and targets.