Retention IQ
2026 methodology edition

State of client reactivation.Evidence before headlines.

This edition publishes the protocol Retention IQ will use before releasing any cross-vertical benchmark. It contains no observed customer averages and makes no performance claim.

Publication status: methodology only. The production ledger currently does not support a qualifying cross-customer report, so planning scenarios remain labeled as planning scenarios.
Release gates

When a number can be called a benchmark.

Every gate below must pass. If one fails, the publication may show methodology or formula-driven scenarios only.

01

Analysis plan and reporting window frozen before results are calculated

02

At least five independent customer organizations in every published cohort

03

Sends, clicks, bookings, cancellations, no-shows, completions, and attribution reconcile

04

Consent and suppression controls active for every included production send

05

Contracts and privacy terms permit de-identified aggregate analysis

06

A reviewer independent of the analysis checks queries, exclusions, labels, and calculations

Canonical funnel

Clicks are not returns. Bookings are not completions.

The future report will preserve every stage and denominator. Cancellations and no-shows cannot be silently removed.

StageDefinitionBillable?
Eligible recipientPassed consent, suppression, tenant, and channel-readiness checksNo
Delivered sendProvider accepted or delivered under the documented definitionNo
ClickValid tracked click after bot and duplicate handlingNo
BookingAppointment or job created and linked under the attribution policyNo
Cancellation / no-showBooking did not become a completed serviceNo
Completed returnSource system verifies the attributed service occurredYes — $75 unless an order form controls

Required metrics

  • Eligible recipients, delivered sends, clicks, bookings, cancellations, no-shows, and completed returns
  • Median send-to-booking and booking-to-completion time
  • Recovered revenue with median and interquartile range, not only totals
  • Fees, disputes, reversals, missing-data rate, and cohort size beside every table

Required claim labels

  • Observed Retention IQ aggregate — every release gate passed
  • Customer-specific result — permissioned and auditable
  • Planning scenario — formula-driven, not an observed result
  • Third-party benchmark — direct source, date, population, and methodology caveat

Want to test the protocol on your own data?

Start with sample data. If the workflow fits, we validate permissions, data quality, consent, attribution, and completion evidence before customer data moves.

15 minutes · no sales pitch Works with your booking platform