Google Ads Enhanced Conversions: Setup and Verification

Enhanced conversions are a measurement layer that hashes the first-party data you collect at conversion time, such as an email address or phone number, with SHA-256 before it leaves the browser and sends that hash alongside your conversion tag to Google Ads. Google compares it against the hash of the account the user was signed into when they clicked your ad; when the two match, the conversion is credited to the campaign even though no cookie could be read. In 2026 the web and leads versions were merged into one module, so the real question is no longer which feature to switch on, but through which path and at what quality you send the data.

What are enhanced conversions and how do they differ from standard conversion tracking?

Standard conversion tracking links a click to a conversion through a cookie or a click ID. Enhanced conversions add a second identity bridge: hashed first-party data. Even when the cookie is deleted, its lifetime is capped by the browser or the user switches devices, a hash match can still recover the conversion. It does not replace the base setup, it sits on top of it; if your tag does not fire correctly, enhanced conversions will not work either. We covered the base setup step by step in Google Ads conversion tracking setup.

• Signal source: standard tracking relies on cookies and click IDs; enhanced conversions use the email, phone number or name and address the user typed in themselves.
• Match logic: nothing travels in plain text. The value is normalized, hashed with SHA-256 and compared against hashes on Google's side; the raw email is never transmitted.
• When it kicks in: only when a conversion fires. Enhanced conversions generate no new demand, they recover conversions that exist but cannot be attributed.
• What stays the same: targeting and bidding logic are untouched. More complete conversion data does enlarge the sample Smart Bidding learns from, and that indirect gain is usually the bigger one.
• What it depends on: an identity field actually being present at conversion time and the user having granted ad personalization consent. If either is missing, the setup can be technically correct and still produce no matches.

How much difference do enhanced conversions actually make?

The gain scales with how much signal you were losing: accounts with high consent rejection and strict browser limits recover more. Industry benchmarks report a single-digit conversion lift on Search as typical, and accounts with severe signal loss can see a larger share come back. Set the right expectation: this is a correction to reported conversions, not new revenue. The curve below shows a sample scenario where the setup goes live in week three and the match rate settles.

Which fields are hashed and how are they normalized?

The accepted fields are email, phone number, and first and last name together with address components such as postal code and country. In practice email delivers the highest match rate; phone and address are complements and contribute little on their own. Normalization before hashing is mandatory: a string that is not normalized produces a different hash, and the match silently drops to zero. Here is the order.

1. Trim leading and trailing whitespace and convert the email address entirely to lowercase.
2. Format the phone number in E.164: a plus sign, the country code, and no spaces or punctuation (for example +14155550123).
3. Lowercase first and last names, strip digits and punctuation, and send the postal code and the two-letter country code as separate fields for address.
4. Hash the normalized string with SHA-256 and transmit the result as lowercase hexadecimal.
5. Carry a deduplication ID for each conversion, such as transaction_id or order_id, so the tag and the API never record the same sale twice.

Google tag, GTM or API: which setup path should you pick?

The choice follows where the data reliably exists. If the email is present on the page at conversion time, the browser-side path is enough. If the email only becomes meaningful in the CRM, for instance a qualified lead or a sale closed over the phone, you need the API or upload path. In practice the most robust configuration is a hybrid where both run and are tied together by a deduplication ID.

• Google tag, automatic collection: the fastest path. The tag tries to read form fields on the conversion page. It is unreliable in single-page apps and wherever the thank-you page exposes no field at all, so always verify from the network request after you enable it.
• GTM with a user-provided data variable: you map fields explicitly through a CSS selector or a dataLayer variable. Control is high, but a CSS selector breaks silently when the page structure changes, so prefer dataLayer wherever you can.
• API and data uploads: you send the conversion from your server or CRM with hashed fields and the click ID. For long sales cycles this is the only realistic path. Note that in 2026 the upload pipeline moved to Data Manager, so legacy API flows need an audit. We covered the wider architecture in server-side tracking.

How do consent signals affect enhanced conversions?

Enhanced conversions depend on consent. Under Consent Mode v2, if ad_user_data is not passed as granted, the hashed payload is dropped server side and no match is even attempted, however active the setting looks. That is why your cookie banner and your measurement setup are really one project. In practice, lifting the consent rate by a few points recovers more conversions in most accounts than any extra tag tweak. For the details see the Consent Mode v2 guide.

1. Open the network tab on your conversion page and confirm that the user-data parameters are populated in the Google Ads request.
2. Switch consent to denied and repeat the test; this time confirm that nothing is sent. Only both tests together are meaningful.
3. Open the diagnostics report 48 to 72 hours after the first send and read the match rate along with any field-level warnings.
4. Compare the Ads conversion count with your own order or CRM count; a gap wider than 15% means a link in the chain is broken.

Is the setup really working, and which thresholds prove it?

A toggle in the on position is not evidence. Three numbers are: the match rate in the diagnostics report, the share of conversions that carry user-provided data, and the gap between the Ads count and your own system. In a healthy setup the match rate stays above 40%, the large majority of conversions carry hashed data, and the system gap sits in the 5-10% band. When that trio degrades, question the data source and the consent flow before you touch the tag.

The real risk in measurement is not data that never arrives, it is data that arrives incomplete: a setup that looks enabled but never matches will train Smart Bidding on the wrong examples for weeks.Ads Sensor Team

Once enhanced conversions are live, the actual work is watching conversion volume and cost every week. You usually notice a falling match rate through a sudden swing in campaign performance, and catching it early requires the data to sit in one place and be comparable. Ads Sensor reads Google Ads and GA4 conversions side by side, flags breaks in volume as anomalies, and shows the reasoning behind every action it proposes. You can sign up for the pre-beta.