Ad platforms learn from the conversions you tell them about. If they only see form fills, they will find you more form fills. Here is how to teach them what a customer looks like.
Why the platforms need your data
Google Ads, Meta and Microsoft Advertising use automated bidding: they show your ads to the people most likely to convert. “Convert” means whatever you report. If that is “submitted a form”, the system learns to find people who submit forms, including people who never buy.
Offline conversions close the gap. When a lead qualifies, a call books an appointment or a deal closes, you send that event back to the platform, matched to the click that started it.
How matching works
- Click IDs. Each platform adds an ID to its ad links: gclid (and gbraid, wbraid) for Google, msclkid for Microsoft, fbclid for Meta. If you kept the ID with the lead, the platform can match the conversion exactly.
- Hashed contact details. Email and phone number, normalised and hashed with SHA-256, let the platform match people who signed in to its services, even without a click ID. Google calls this enhanced conversions; Meta uses it in its Conversions API.
Both methods depend on normalising details the same way before hashing: lower-case emails, phone numbers in international format.
Getting started
- Keep click IDs and the first and last touch with every lead (see tracking lead sources in forms).
- Decide which later events matter: a qualified lead, a booked call, a sale over a certain value.
- Create a conversion action for each in the ad platform.
- Send each event once, as soon as you know about it.
- Give the platform a few weeks to learn before judging the change.
In Metric Sensei
Map an event, such as a Salesforce opportunity won, a CallRail call or a Shopify order, to a conversion action, and Metric Sensei sends it to Google Ads, Meta or Microsoft Advertising, matched on click IDs and hashed contact details. Each event goes to each conversion action once. See Attribution.