Set up Google Ads offline conversions by linking a defined business outcome to the original lead, preserving the supported identifiers and validating what the import actually records. Start with one outcome, such as an accepted appointment or completed sale. A lead entering a spreadsheet is not automatically the same event as a customer completing the work.
The difficult part is usually the handover between the website, the sales record and the person maintaining advertising measurement. Agree which record owns the outcome and when it becomes final enough to import. Then build the data mapping and exception process around that decision, rather than exporting every contact and hoping the account interprets it correctly.
Which identifiers and events belong together?
Offline conversion imports can measure outcomes that happen after an advertising click or call. Google Ads calls its click identifier the Google Click ID, or GCLID. The offline import overview describes a click-based route and a separate call-based route. Choose the route that matches how the lead arrived and retain that context in the implementation record.
The documented GCLID workflow saves the click identifier alongside the collected lead information. The later offline conversion submission includes the GCLID, conversion type and when the conversion happened. Therefore, a useful test follows one identifiable lead through the complete business process, instead of checking the website and the import file as unrelated technical tasks.
Enhanced conversions for leads can add user-provided data to the identifiers already imported. Decide which supported method the account will use before defining the export. Google’s current offline-import overview recommends the Data Manager API for current and future offline-conversion uploads. Google’s notice says offline conversion import and enhanced-conversion-for-leads uploads move to the Data Manager API from 15 June 2026. It says developer tokens without a request between January and June 2026 will not be allowlisted for legacy access. Record the chosen upload route instead of reusing an older Google Ads API specification.
Keep a business event identifier in the internal worksheet even when it is not an upload field for the chosen method. Its purpose is to help the business explain which outcome a row represents. Do not invent a GCLID or substitute a customer reference for an advertising identifier when the required match information is missing.
What makes the conversion definition usable?
Offline conversion setup asks the advertiser to choose a category related to the measured action. Offline conversion actions can be configured as primary or secondary. The setup guide makes both decisions part of configuration. Record the business definition next to the chosen category and role so the account owner can explain what the imported event means.
For example, define whether “qualified enquiry” means a salesperson has checked a requirement, the customer has booked an appointment or a formal opportunity has been accepted. Pick the state that matches the organisation’s actual process. If staff use the term differently, resolve that disagreement before connecting the measurement to bidding decisions.
Offline conversion setup includes selecting source fields and any additional fields to import. Map the source field, its meaning, its format and the person responsible for correcting it. Keep event time distinct from export time. A later batch upload should not silently replace the time at which the business event occurred.
An offline conversion action needs a connected data source before it can receive and record data. Creating the action is therefore a configuration milestone, not evidence that the complete measurement route works. The acceptance record should show the connection, the selected mapping and an observed result from the intended process.
| Worksheet field | Business meaning | Validation question |
|---|---|---|
| Internal lead reference | Which enquiry is being followed? | Can staff find the source record? |
| Supported advertising identifier | Which ad interaction may be matched? | Was it captured through the chosen route? |
| Outcome | What actually happened? | Does it meet the written conversion definition? |
| Event time | When did the outcome happen? | Is the source time and zone understood? |
| Permission status | Approved handling for this data | Has the responsible owner reviewed it? |
| Import observation | What the destination accepted or rejected | Is there evidence beyond an exported file? |
| Exception owner | Who resolves a missing or rejected field? | Is a specific next action assigned? |
How should consent and data handling enter the workflow?
Ask the responsible data owner to approve the collection, sharing and retention approach before enabling the proposed flow. Explain the actual fields and destinations in that review. A generic statement that the data is “for marketing” is insufficient for the implementer to know what may be sent through this particular route.
Keep personal information out of demonstration screenshots and shared issue reports. Use fictional records for design reviews, then use an authorised validation method for the real integration. Record where sensitive diagnostic detail may be inspected without copying entire customer records into an ordinary campaign brief.
Enhanced-conversion diagnostics can identify missing or incorrectly formatted user-provided data. Use the relevant diagnostic view as part of investigation, while also checking the source record and mapping. A formatting correction should preserve the underlying meaning; it should not become a reason to manufacture a missing customer attribute.
The conversion tracking guide covers the wider event and consent design. Analytics event planning helps separate website activity from business outcomes. This workflow owns the later outcome and its traceable import.
Original lead-to-outcome import rehearsal
Use this hypothetical exercise for a consultancy that receives enquiries and later confirms appointments. The original artefact is a four-record import rehearsal. Each record has an expected treatment before any upload is attempted, making it possible to identify a process failure as well as a technical rejection.
Record A: completed appointment. The lead has the supported identifier, a documented completed appointment and an approved handling status. Trace its source fields to the planned import mapping. Record the intended conversion action and event time. Compare the observed import result with the expected single outcome, retaining the internal reference used to make that comparison.
Record B: enquiry only. The person has asked a question but no appointment has been confirmed. The expected treatment is to remain outside this completed-appointment batch. If the export includes it, correct the business-state filter. A technically accepted row would still be the wrong measurement for this event definition.
Record C: missing match information. The appointment exists, but the selected route lacks a required identifier or field. Hold the row for the assigned owner. Record whether an approved alternative method exists; do not invent identifiers or copy another lead’s data to make the row fit.
Record D: repeated export candidate. The same business outcome appears in two source views. Compare the internal references and inspect how the chosen import method handles repeat submissions. Define the controlled retry procedure from the actual method’s documentation and observed behaviour before automating it. This test should settle ownership of duplicate candidates, not assume that every destination deduplicates every field combination.
Finish the rehearsal by reconciling the intended batch, submitted rows, accepted observations and unresolved exceptions. Counts alone are insufficient: explain which record changed state and why. Give each unresolved row an owner and preserve the original event evidence.
For broader campaign planning, see Google Ads. To prepare an integration brief, get started with the outcome definition, source fields and the four-record rehearsal results.
By Spout Team. AI-assisted drafting by Codex using the linked primary documentation.