Marketing may call a form submission a qualified lead while sales uses the same phrase for a prospect who completed a discovery call. Sending those inconsistent labels into reporting or bidding makes the system optimize against a moving definition.
A lead-stage conversion map resolves the meaning before the integration. It connects the event the business cares about to the record that proves it happened and the advertising signal, if any, that should represent it.
Start with the real sales process
Interview the people who handle inquiries and inspect representative records. Write the steps customers actually take: inquiry, validation, contact, qualification, appointment, proposal, agreement, fulfillment, or payment as relevant.
Do not copy a generic funnel if it does not match the business. A local service appointment and an enterprise procurement process need different milestones and time horizons.
HubSpot's lifecycle-stage documentation describes contact and company stages, customization, and their relationship to other status fields. A CRM's defaults are a starting vocabulary, not evidence that your team consistently uses those definitions.
Distinguish the entity from the event
A contact can be associated with several opportunities over time. A company can contain several contacts. A current stage is a state, while entering that stage is an event with a time and supporting evidence.
Choose the entity for each measurement. Unique qualified contacts, qualified buying opportunities, and completed jobs answer different questions. Avoid counting all three as equivalent conversions in the same performance metric.
Retain event history where the integration requires it. Overwriting a contact's current status should not erase when qualification occurred or create another conversion every time a record is synchronized.
Use a mapping table with acceptance evidence
| Business event | Evidence | Time to preserve | Owner |
|---|---|---|---|
| Inquiry received | Valid form or call record | Original receipt | Marketing operations |
| Qualified opportunity | Agreed criteria and accepted record | Qualification event | Sales |
| Appointment attended | Completed meeting or visit | Attendance | Sales or service team |
| Agreement completed | Accepted deal record | Actual completion | Sales operations |
| Revenue adjusted | Cancellation or financial adjustment | Adjustment and original reference | Finance or operations |
Adapt the table to the business and identify any stages that are informational only. A submitted form does not need to become a bidding signal simply because it is easy to collect.
For each stage, add the exact criteria and a counterexample. “In our service area and requesting an offered service” is clearer than “good lead.” A person who never answered remains unknown unless another criterion establishes disqualification.
Map stages to platform goals deliberately
Google's qualified- and converted-lead documentation describes categories for deeper events defined by the advertiser's process. Choose the category and action that match the event; do not use “converted” as shorthand for revenue if the chosen milestone is earlier.
Record the conversion-action identifier, category, counting rule, value rule, and campaign-goal use. Inspect whether multiple actions represent the same business outcome and could lead to overlapping reporting or incentives.
Keep source events and imported advertising events linked. That allows the team to investigate missing or duplicate signals without exposing unnecessary customer information in broad reports.
Define timestamps and delay explicitly
Preserve when the business event occurred separately from when the integration received or uploaded it. A batch imported on Friday may contain qualifications from several earlier days.
Document the timezone and expected delay at each stage. If a salesperson updates a week's records at once, the update time may not represent the actual sales event. Decide how to handle that limitation before building daily optimization rules.
The offline import QA guide covers transport and processing checks. A successful upload does not prove that the underlying stage definition or timestamp is correct.
Handle repeated stages and reversals
Agree what happens when an opportunity reopens, a meeting is rescheduled, or a contract is canceled. Determine whether the event is new, an update, or a correction under the business and platform's supported behavior.
Use stable event references and the integration's documented deduplication mechanism. Do not invent a universal rule that a contact may generate only one conversion forever, or that every status change should create a new one.
Keep invalid or unsupported corrections visible for manual resolution. Silently deleting historical events can break reconciliation, while repeatedly uploading them can inflate counts.
Choose values that match the event's meaning
Realized revenue, expected deal value, and a relative quality score are different value systems. State which one is being used and why. If expected value depends on a close probability, record its source and update method.
Do not assign today's closed-won revenue to every earlier stage as though those amounts were independent realized sales. If several stages carry values for separate purposes, explain how reports and bidding avoid double interpretation.
Test the mapping on a small authorized sample before expanding. Reconcile source counts, eligible events, uploaded events, processing results, and eventual reporting under the platform's timing and matching rules.
Give the map an owner and a change process
Qualification criteria change as the business enters markets or revises its offer. Record versions and effective dates so historical cohorts remain interpretable. A new definition can change reported quality even when the acquired audience stays similar.
Review examples with marketing, sales, and operations until they classify the same records consistently. Use the long-cycle reporting guide to preserve cohort maturity once the map is live.
A stage map is complete when the team can point to an event, prove its meaning, trace its timestamp and value, and explain how it affects advertising decisions. The integration becomes much easier to trust once that business contract is explicit.