Google Ads Enhanced Conversions After the June 2026 Merge
Google announced one switch for web and leads enhanced conversions from June 2026, and leads uploads moved to the Data Manager API. Here is how Google combines data from each source and how to prove your setup works.
Google announced that from June 2026, Google Ads would stop asking which kind of enhanced conversions you run, and trade press reports it has rolled out. Enhanced conversions for web and enhanced conversions for leads become a single feature behind one on/off switch, and the screen where you picked the Google tag, Google Tag Manager or the Google Ads API as your method goes away. Google now accepts hashed first-party data from your website tag, from Data Manager and from API connections at the same time, and combines them for each conversion.
Google said accounts already using the feature would migrate automatically, provided they had accepted the customer data terms, so for most advertisers nothing visibly broke. That is the reason to audit it now. One green toggle now covers several data paths, and it tells you nothing about whether the tag, your CRM data uploads or a server feed actually reach Google Ads attribution in a usable form. A second change landed the same month. Since June 15, 2026, offline conversion and leads uploads must use the Data Manager API, and the Google Ads API blocks new ones. If you rely on Smart Bidding through the holiday season, both changes deserve an hour of checking before Black Friday.
What Changed When Google Merged Enhanced Conversions
Enhanced conversions for leads launched in March 2022 as a separate product. The web version captures an email, phone number or address on the conversion page, hashes it with SHA256 and matches it against signed-in Google accounts. The leads version captures the same kind of data when someone submits a form, then waits for you to import the offline outcome, such as a qualified lead or a closed deal, from your CRM with matching identifiers.
Google’s help page on the updates to your enhanced conversions settings describes the change in two phases. In April 2026, Google Ads started accepting user-provided data from website tags, Data Manager and API connections simultaneously. The June 2026 phase removes the web and leads labels from the interface, along with the method selector, and leaves one status and one switch.
The method selector mattered more than it looked. Google’s own setup guide for the tag still warns that Google may not process data sent through a method other than the one you selected. In practice the dropdown acted as a filter, so an account set to the Google tag could quietly ignore an API feed. The merge removes that filter, which is the real improvement. A tag blocked by a browser or a consent choice is no longer the only way user data can reach a conversion.
Several things stay exactly as they were. Matching still depends on signed-in Google accounts. You still have to accept the Google Ads Data Processing Terms before Google processes anything. Consent still gates the data. And the quality of what you send is still your problem. A merged switch will not make a scraped or malformed email address match.
Where the Single Setting Lives Now
You can turn enhanced conversions on in two places, and the account-level setting is the default for every conversion action.
- Account level: Goals, then Settings, then the Customer data use panel. Check the enhanced conversions box, select Agree and save, and confirm the data terms on the second screen. This applies to every conversion action unless an action overrides it.
- Conversion action level: the same Customer data use panel appears when you create or edit a conversion action, and you can opt a single action out at any time.
Opting an action out makes sense when it never carries customer data, such as a page view goal or a scroll event. For purchase and lead actions, leave it on.
The same panel holds a second checkbox that deserves a deliberate decision: conversion-based customer lists. When it is on, Google creates an audience segment for each conversion goal that has enhanced conversions active, and Smart Bidding uses Customer Match lists automatically. Trade press, citing Google’s emails to advertisers, reported that Google turned this setting on automatically from June 17, 2026, for accounts already running both enhanced conversions and Customer Match, with data processing starting August 18. If your account was in that group and nobody opted out, Audience Manager now holds lists built from your converters. Check that your privacy notice covers that use.
Shopify merchants have one more switch. The Google and YouTube app keeps its own enhanced conversions setting under Settings, then Google Ads settings. Make sure it agrees with what the Google Ads account shows, so nobody assumes the other side handled it.
How Google Combines One Conversion From Several Sources
Once several paths can send data for the same purchase, the obvious worry is double counting. Google answers it in the Data Manager API documentation for Google Ads multi-source conversions, and the join key is the transaction ID. When an uploaded event reaches the same conversion action as a tag conversion:
- If the transaction IDs match, the uploaded conversion value and currency replace the tag’s values.
- If the tag captured no user data, Google adds the uploaded user data to that conversion. If the tag already captured some, Google ignores the uploaded user data.
- Google ignores other fields on a matched upload, such as the GCLID.
- If no tag conversion shares the transaction ID, the upload becomes a new conversion and Google tries to attribute it to an ad click.
Two timing rules apply on top. Each conversion action starts with a 14-day trial for multi-source data, during which value updates are disabled and new uploaded conversions show in reports without feeding bidding. Google also recommends uploading within 24 hours of the original tag event. Google’s help center still labels multi-source conversions as beta, and they only work on website conversion actions set up with code through the Google tag or Google Tag Manager, not on imported GA4 conversions or URL-based actions.
Three practical consequences follow. First, a transaction ID on the tag is no longer optional. Without it, every server upload looks like a fresh conversion. Second, both sides must carry the identical string. An order name on the tag and a numeric order ID in the upload will never match. Third, a server feed cannot repair a tag that captures the wrong data, because user data the tag already sent wins. If automatic detection is scraping a newsletter field instead of the checkout email, fix the tag.
For a code-based tag setup, the pattern looks like this, with normalization done before the values reach the page:
// Set the customer's details before the conversion event fires.
// Google hashes plain values on secure (HTTPS) pages; you can also send
// pre-hashed sha256_ fields. Leave out any field the site does not collect.
gtag('set', 'user_data', {
email: 'jane.doe@example.com', // trimmed and lowercased
phone_number: '+12125550123', // E.164 format with the + prefix
address: {
first_name: 'jane',
last_name: 'doe',
postal_code: '10001',
country: 'US' // ISO 3166-1 alpha-2 code
}
});
// Fire the conversion with the same order ID your backend uploads later.
gtag('event', 'conversion', {
send_to: 'AW-XXXXXXXXX/CONVERSION_LABEL',
value: 129.00,
currency: 'USD',
transaction_id: 'ORDER-104233' // the deduplication key
});In Google Tag Manager, the equivalent is the Transaction ID field in the Google Ads Conversion Tracking tag, fed by a data layer variable, plus the user data mapped into that tag. Preview mode shows that data as the Google Ads enhanced conversions object. Expect reported conversions in Google Ads to rise once more data matches, which widens the usual gap between Google Ads, Meta and Shopify revenue figures. That gap is structural. Do not treat the jump as a tracking error.
What Data Manager and GA4 Can Feed Into It
Data Manager sits under Tools, then Data manager. It works with three concepts: a data source (a connected product, file or third-party integration), a connection (the table or file you import) and a destination (where the data is used). Google’s table of supported sources lists 18 direct connectors, and the gaps matter more than the length of the list:
- Shopify supports Customer Match but not offline conversion import.
- Salesforce supports offline conversion import but not Customer Match.
- BigQuery, HubSpot, Klaviyo, Snowflake, Zoho CRM, Google Sheets, SFTP and HTTPS support both.
So Google’s table gives a Shopify store no native Data Manager route for importing its orders as offline conversions. Online orders still reach Google Ads through the tag, through the API’s multi-source path, or through a warehouse such as BigQuery that Data Manager can read. Lead generation businesses have the cleaner route, since their CRM data is exactly what the leads use case was built for.
One more change lands on November 16, 2026. Google told advertisers that Data Manager will stop requiring SHA-256 hashed customer data and will accept encrypted data by default, decrypting and matching it inside a trusted execution environment. The email names enhanced conversions and covers every Data Manager route, from CRM connections and file uploads to the Data Manager API. Google’s formatting guide still describes SHA-256 for emails and phone numbers, so ask whoever built your uploads whether anything needs to change before you touch a working pipeline.
GA4 is the other path. With user-provided data collection turned on under Admin, then Data collection, and the property linked to Google Ads, GA4 sends hashed data that supports enhanced conversions and Customer Match. This is the path that matters when your Google Ads conversions are imported GA4 key events rather than native Google Ads tags. If those key events came out of the migration in my GA4 key events guide, check this setting before assuming the merged switch covers them. Google says model and attribution improvements can take up to a month to show, and properties in the Health industry category cannot use the feature. Conversion-based customer lists follow the same route: with the GA4 path opted in, those lists fill with GA4 key events.
One setting can silently cancel all of this. The Google tag has an Allow user-provided data capabilities setting, reachable from Google Ads under Tools, then Data manager, then Google tag, and from the Settings screen in Google Tag Manager. When it is off, no product that uses that tag receives user-provided data from it.
The June 15 Cutoff for Offline and Leads Uploads
Since June 15, 2026, offline conversion imports and enhanced conversions for leads uploads go through the Data Manager API and are blocked in the Google Ads API. Google’s help center and its developer blog give slightly different windows for who kept legacy access (January to June 2026 in one, December 2025 to May 2026 in the other), but the rule is the same: tokens with no recent offline conversion imports were not allowlisted. Google’s developer blog says those tokens get the error CUSTOMER_NOT_ALLOWLISTED_FOR_THIS_FEATURE from ConversionUploadService.UploadClickConversions, and existing adopters can keep using the Google Ads API while they integrate with the Data Manager API.
The migration is more than an endpoint swap. The differences that trip up engineers:
- The Data Manager API needs new credentials with its own scope, and it does not use a developer token at all.
- Quotas sit on the Google Cloud project. Ingestion allows 100,000 requests a day and 300 a minute, with up to 2,000 events per request and up to 10 user identifiers per event.
- The Data Manager API fails fast. One bad record rejects the whole request, while the Google Ads API accepted partial failures. Run requests with validateOnly set to true first.
- The operating account must be the account that owns the conversion action. The Data Manager API does not allow the parent and child accounts that the Google Ads API accepted.
- To restate a value or add user data, send the event again with the same transaction ID. Retractions are not supported.
Keep sending GCLIDs whenever you have them. Google recommends them in all cases and requires them when no tag collected user-provided data for the lead. If a connector vendor handles your uploads, ask them directly whether they have moved to the Data Manager API, and get the date in writing.
If you are still on plain offline conversion import and want to upgrade to the leads version, Google’s guidance is to run the new action alongside the old one for one to two conversion cycles or four weeks, whichever is longer, before making it primary.
How to Read the Enhanced Conversions Diagnostics
The enhanced conversions for web tag diagnostics report lives under Goals, then Summary, on the Diagnostics tab. It rates each setup as Excellent, Good, Needs attention or No recent data. The last one means Google recorded nothing in the last seven days. The leads version swaps No recent data for an Urgent status, used when tags are not firing at all or every ping arrives without user data.
Two numbers do most of the work, and they diagnose different failures.
Coverage is the share of conversion events that arrived with user-provided data attached. It defaults to the last 7 days and can switch to 30. Low coverage means the data is missing at the source: some pages do not send it, or selectors break on mobile or in certain browsers.
Match rate is how much of the data you sent matches signed-in Google accounts. Google reports it in buckets: INSUFFICIENT_VOLUME below 20 valid pings, NO_MATCHES, LOW from 0 to 15%, and HIGH above 15%. Healthy coverage with a low match rate points to normalization or hashing problems, or to sending only one identifier where three were available.
When a single conversion looks wrong, inspect the em parameter on the conversion hit. A value of tv.1~em. with nothing after it means the field was empty. The codes e0 to e3 mean, in order, failed formatting, an unsupported legacy browser, a failed SHA256 hash and a page not served over HTTPS. A valid hash runs well past ten characters.
Timing explains most false alarms. The diagnostics report appears about 72 hours after implementation, alerts need at least 20 conversions in the past seven days, and the impact figures need up to 30 days of training before they show. In the report itself, Google recommends adding in-page code on top of automatic detection. The same field-by-field discipline used for raising match quality in a server-side GTM container applies here too.
An Enhanced Conversions Audit Checklist
Run this once per account, then again after any checkout, form or consent change:
- In Goals, then Settings, confirm enhanced conversions is on and the data terms are accepted. List every conversion action opted out at action level and write down why.
- Confirm each primary purchase and lead action shows enhanced conversions as active. On Shopify, match the Google and YouTube app setting to the account.
- Open the Google tag settings and confirm Allow user-provided data capabilities is on.
- Fire a test conversion with Tag Assistant or the browser network tab, and check that the em parameter carries a full hash, with no empty value and no e0 to e3 code.
- Confirm the tag sends a transaction ID, and that any upload uses the identical value.
- Test the consent path. When ad_user_data is denied, no hashed data should leave the page. When it is granted, it should.
- Read coverage and match rate per action for 7 and 30 days, and fix coverage before chasing match rate.
- For leads and offline data, confirm uploads run through Data Manager or the Data Manager API, target the account that owns the action, and include GCLIDs where available.
- Check Audience Manager for conversion-based customer lists and decide whether you want them.
- Annotate the date of every fix in your reporting, so a rise in Google Ads conversions is not mistaken for a sales trend when you compare it with Shopify orders and GA4 channel reporting for a Shopify store.
The June merge simplified the switch. What feeds it is still yours to check, and the diagnostics tab shows which of your data paths is actually doing the work.
