Choose one outcome stage to establish first
A submitted form, staff-qualified inquiry, appointment, and completed sale are different stages. Define the first imported stage in observable terms and identify who records it. For example, qualification might require an offered service, a usable contact route, and a deliverable job area. Do not label every incoming record qualified to increase the upload count. The signal needs to describe a consistent business decision rather than a marketing aspiration.
Source checks: Configure the Google tag for enhanced conversions for leads · About primary and secondary conversion actions
Check the currently supported integration path
Google's current enhanced-conversions documentation includes Data Manager and notes 2026 changes to legacy API upload support. Follow the process actually supported for your account and connector instead of copying an old API recipe. Review the relevant data-use and consent requirements with the implementation owner. Hashing or matching support is not permission to collect arbitrary personal information or send it through general analytics parameters.
Design a small record schema
Include a stable internal outcome key, stage, outcome date and timezone, supported matching identifiers, and a value only when its meaning is defined. Record inquiry creation separately from later qualification or sale. Keep sensitive data in the restricted system and expose only the fields the supported import requires. Maintain an audit field showing whether a record is a test, pending review, or ready for import.
Prevent duplicate and conflicting outcomes
Decide whether a later stage updates an earlier record or creates a different conversion action according to the supported setup. Use stable identifiers rather than generating a new key on every export. Re-exporting the same qualified outcome should not become another new customer. Document corrections and canceled work so the team can use supported adjustment procedures when applicable. A spreadsheet copy without identity rules can silently multiply outcomes.
Validate a limited sample before broad use
Check stage labels, dates, timezone offsets, required identifiers, and any reported diagnostics or errors. Reconcile accepted import records with the source file without treating acceptance as proof that every record matched an ad or represented completed revenue. Inspect action optimization settings so the imported stage is used as intended. Keep a supported test process and avoid feeding fictional conversions into live optimization as customer results.
Review outcome quality after connection
Compare uploaded, accepted, attributed, and business-confirmed outcomes as separate counts where the tools expose them. Investigate duplicate keys, late stage updates, and changes in staff qualification. Schedule a practical data review rather than assuming the connector makes the source records reliable. The useful result is a repeatable, appropriately controlled transfer of a defined business outcome, with a clear owner and documented limits.
Worked example: illustrative figures
For example, a source sheet has 25 rows, but 5 are repeated exports of the same qualified outcomes. Stable outcome keys reduce it to 20 distinct records. Three are still pending staff review, leaving 17 ready records. An import summary later accepts 16 and flags 1 error. The owner reports 16 accepted records, not 25 new customers; ad attribution and completed sales remain separate questions. These figures teach the preparation and reconciliation process without guaranteeing platform match rates or prescribing an unsupported upload method.
Prepare outcomes before connecting the import
25 source rows
Check repeated exports and internal keys.
20 distinct
Remove five duplicate representations.
17 ready
Three outcomes still need staff review.
16 accepted
Resolve one error; check attribution separately.
Illustrative 25-row reconciliation. Accepted records are not automatically new customers or attributed sales.
Your practical checklist
- Define one observable outcome stage and owner.
- Use the account's current supported import path.
- Retain outcome date and timezone explicitly.
- Use stable non-personal outcome keys.
- Separate tests, duplicates, and pending review records.
- Reconcile source, accepted, and attributed counts.
Your next action
Prepare ten existing outcome records with a stable key, consistent stage, outcome time, and clearly documented import readiness.
Questions about this task
Can I import every form as a qualified lead?
Only if every form genuinely meets the written qualification definition, which is uncommon. Submitted and qualified stages should remain distinct when staff review is required.
Does an accepted upload prove a customer came from ads?
No. Acceptance, attribution, qualification, and sale are different stages. Read the diagnostics and reconcile with the business records.
Sources and editorial notes
Prepared with AI-assisted research. Official sources are linked below; worked examples and checklists are Getting Found Local recommendations, not measured client outcomes.
Sources checked October 13, 2026. Platform interfaces and local resources may change; follow the current instructions on the linked official page.
- Configure the Google tag for enhanced conversions for leads ↗
- About primary and secondary conversion actions ↗
How we research, illustrate, and update these guides · Suggest a correction