The Jargon Barrier in Audience Targeting

If you have spent any time researching intent-based advertising, you have encountered the same cluster of terms: Device ID, hashed email, MD5, custom audience, first-party data, match rate. These terms often appear together without enough context to make the overall picture clear.

This guide explains each concept in plain language and walks through exactly how a search query performed on a phone turns into a verified targeting record inside a Meta campaign.

What Is a Device ID?

Every internet-connected devicecarries a unique identifier. On Apple devices, this is the Identifier forAdvertisers (IDFA). On Android devices, it is the Google Advertising ID (GAID).

When someone searches forsomething on their device, that search can be associated with the device'sidentifier. The connection creates a record: this device performed this queryat this time.

A Device ID by itself is not aname, an email address, or any other personally identifiable piece ofinformation. The linkage to a real person occurs only when the Device ID ismatched against a database of known user profiles. Device IDs alone are consideredpseudonymous rather than personally identifiable under most privacy frameworks.¹

What Is Hashing? What Is MD5?

Hashing is a one-waymathematical transformation that converts any string of text into afixed-length output. When you apply the MD5 algorithm to an email address, theresult is a 32-character string that cannot be reversed.

This is what makes hashing safefor advertising contexts: you can share a hashed email list with a platformwithout sharing the underlying contact data. The platform hashes its own useremail records using the same algorithm and compares them to your list. MD5 isthe algorithm used by Meta, Google, TikTok, and LinkedIn for their customermatch products.

What Is a Custom Audience?

A custom audience is a targetingsegment you build from your own data rather than the platform's modeledinferences. Instead of asking the platform to find people who might beinterested in your product, you hand it a list of specific people you want to reach.

Custom audiences bypass theplatform's audience selection process entirely. The platform's role becomesdelivery, not discovery.

How the Full Workflow Connects: Step by Step

Step 1: The Search Happens

A person searches "cheapestauto insurance Texas" on their iPhone. That search is associated withtheir IDFA.

Step 2: Intent Classification

The search query is classifiedagainst the IAB content taxonomy. This search maps to In-Market > Financial> Insurance > Auto.

Step 3: Identity Resolution

The device identifier is runagainst an identity resolution database that links device IDs to known emailaddresses. Well-sourced datasets targeting active searchers typically achieveidentity match rates above 80 percent.²

Step 4: Optional Enrichment

If enriched contact data isrequired, additional fields can be appended: full name, mailing address, phonenumber, and other attributes. This costs more per record than a basic hashedemail match.

Step 5: Hashing and File Preparation

Email addresses are hashed withMD5 before upload. Most platforms accept either raw emails or pre-hashed MD5strings. Raw contact data does not persist on the platform beyond the matchingoperation.

Step 6: Platform Upload and Matching

You upload the file to MetaCustom Audiences, Google Customer Match, TikTok Custom Audience, or LinkedInMatched Audiences. Platform match rates for well-sourced intent lists typicallyrange from 60 to 90 percent.³

Step 7: Audience Activation

Your custom audience is live.You run ads exclusively to users who searched for your category within yourspecified lookback window. The audience can be refreshed daily.

What This Data Cannot Be Used For

•      These records are not opt-in email marketing contacts.They cannot be used for email outreach without separate consent.

•      They cannot be used for SMS or text message marketing.Unsolicited SMS to non-opt-in records constitutes a TCPA violation.

•      They cannot be sold or transferred to third partiesoutside of the specific campaign use case.

Privacy and Compliance Considerations

Under GDPR in the EU, processingbehavioral data requires a lawful basis. Under CCPA in California, consumershave the right to opt out of the sale of their personal information.´ Buyers should confirm that their data providersources records from compliant relationships.

Why Freshness Is Non-Negotiable

The practical value of a DeviceID or hashed email tied to a search intent signal decays rapidly. Effectiveintent targeting requires data refreshed on a 24-hour cycle. Campaigns built onstale intent data produce results that are functionally indistinguishable frombroad demographic targeting.

Frequently Asked Questions

Can I send email to Device ID matched records?

No. These records are for adplatform custom audience use only. Using them for email outreach requiresseparate opt-in consent from each contact.

What match rate should I expect on Meta?

Well-sourced intent liststypically achieve 70 to 90 percent match rates on Meta due to its largeauthenticated user base.

How often should I refresh my custom audience?

Daily refresh is recommended.Intent decays within 24 to 72 hours so stale lists lose their targetingadvantage quickly.

Sources

1. Federal Trade Commission,Mobile Advertising Identifiers and Consumer Privacy.ftc.gov/reports/mobile-privacy-disclosures

2. LiveRamp, Identity ResolutionBenchmark Report 2025. liveramp.com/resources/identity-resolution-benchmarks

3. Meta for Business, CustomAudiences: Match Rates and Best Practices. business.meta.com/help/audiences

4. European Data Protection Board, Guidelines on Personal Data in Advertising. edpb.europa.eu/our-work-tools/our-documents/guidelines