Skip to content

Blog

How to test lead routing before a flow goes live

The time to find an unclear handoff is before the first real lead enters it—not after sales asks where the record went.

2 min read

A form submission does not become pipeline because the form worked. It still needs treatment, ownership and a safe path for exceptions. Those decisions are easy to leave implicit when marketing configures the source and sales inherits the result.

A governed lead flow makes the handoff inspectable before launch. It connects the source to the selected treatment, applies the ownership rules and lets the team simulate an example without publishing the flow.

Choose the treatment before the route

RevTech separates a Basic Flow from an AI Flow. A Basic Flow handles association and routing. An AI Flow can add enrichment, research, scoring and messaging. The choice should follow the work the motion needs, not a preference for more automation.

Because the flow type defines the operating path, make the choice before building around it. Start with the smallest treatment that can satisfy the response standard and preserve the appropriate human decision.

Lead-flow treatment choice
Basic flowAI flow
Associate the source and route ownership.Add approved enrichment or research before routing.
Use when the required context already exists.Use when additional context changes prioritization or treatment.
Keep the handoff predictable and narrow.Define review and fallback for every prepared output.

Make the owner rules explicit

Ownership should resolve from rules the team can read: geography, segment, existing account relationship, campaign, product interest or another approved field. The fallback matters as much as the happy path. A record that cannot be assigned should enter a visible exception queue rather than disappear or inherit an accidental owner.

  • The form and campaign associations are correct.
  • Existing-account ownership is respected where required.
  • New-record routing uses current territory or team rules.
  • A named fallback owns unresolved assignments.
  • Duplicate and missing-data behavior is defined.
  • The response expectation has an accountable owner.

Use simulation as an acceptance test

A simulation should answer more than whether the configuration saves. Run representative examples through the route and inspect the treatment, owner, prepared work and exception behavior. Include at least one case that should not follow the default path.

  • A net-new target account with complete information.
  • An existing customer or open opportunity.
  • A record with a missing or ambiguous routing field.
  • A duplicate identity or conflicting account match.
  • A source that should be associated but not sent to sales.

Publish only when both teams can explain the result

Marketing should be able to explain why the source enters the flow. Sales or RevOps should be able to explain why the treatment and owner are correct. When both sides can read the simulation result and agree, publishing becomes an operating decision rather than a leap of faith.

Re-run the test when the form, campaign association, ownership rule or treatment changes. A flow that passed under yesterday’s routing logic has not automatically passed under today’s.

Frequently asked questions

Test the source association, treatment type, owner resolution, fallback, duplicate behavior, missing-data behavior and the work prepared for the receiving team.
A Basic Flow handles association and routing. An AI Flow can add approved enrichment, research, scoring and messaging before or around the handoff.
It should enter a visible exception path with a named fallback owner rather than disappearing or receiving an accidental default owner.
Repeat it whenever the source, campaign association, treatment, routing rules, permissions or required data materially changes.

Put the framework to work

Demo

Try the demo.

See agents carry the repeatable work of GTM across sales, marketing, customer success, and RevOps. Every action prepared, reviewed, and recorded. Fictional data, real product.

Explore the demo