Knock

Marketing automation platforms price on seats and contacts while the real cost driver is the number of live journeys, integrations, and approval gates an agency maintains. Agencies that scope retainers against journey count and exception handling protect margin; those that sell unlimited automation on a flat fee absorb every new trigger a client requests. The gap between those two postures compounds quietly across a book of business.

Journey debt is the accumulated maintenance obligation inside every live workflow: broken integrations, stale scoring rules, undocumented branches, and approvals that no longer match how the client actually sells. Unlike a bad campaign, journey debt does not surface in a single report; it surfaces as slower delivery, more exception handling, and margin erosion across the retainer. The framework says price the deployment against the journeys you will maintain, not the journeys you launch. A client with 12 active journeys and 4 integrations carries roughly triple the maintenance load of one with 4 journeys, even when both pay the same build fee. The category description makes this explicit: commercial terms should follow the actual number of journeys, integrations, approvals, and maintenance obligations. Forrester's October 2026 finding that governance and security teams operate without a shared working model is the same failure pattern inside an agency, where the person who builds the workflow is rarely the person who inherits its upkeep.

Marketing automation platforms are sold on the happy path: lead captured, score rises, sequence fires, sales gets a clean handoff. The billable work lives on the exception path. Duplicate records, a lead who replies mid-nurture, a webhook that fails at 2am, a WhatsApp opt-out that never syncs back to the CRM. Agencies that map who owns each exception before go-live can price maintenance honestly; agencies that do not absorb the cost silently for months. The framework asks three questions per workflow: what breaks, who is paged, and what the client sees when it does. A concrete trigger came from OpenAI's October 6, 2026 Ironclad collaboration, where agents are trained to apply company-specific business rules inside specialized software. That is the same discipline agencies need for client journeys: encode the rule, name the owner, log the failure. Platforms such as ActiveCampaign, Brevo, and Manychat all expose the triggers; none of them decide who answers when one misfires.

Handoff Baseline Transfer treats the measurable operating baseline, not the platform login, as the deliverable that changes hands at the end of an engagement. A marketing automation build is only as valuable as the numbers a client can reproduce without the original builder: open and reply rates by journey, lead-to-opportunity conversion, list hygiene thresholds, and the named owner of each exception path. When those figures live in a spreadsheet the client cannot regenerate, the retainer is priced on dependency rather than on results. Agencies that document the baseline before launch can hand a client a working system and still charge for optimization, because the next phase is measured against a known starting point. The same discipline protects the agency when a client brings in a second vendor: the baseline is the shared reference, so scope disputes resolve against data instead of memory.

Count the journeys, integrations, approval gates, and monthly maintenance hours first, then set the retainer against that count instead of the number of platform logins.

Count every live journey, integration, and approval gate before quoting, and price the maintenance obligation rather than the platform seat.