Skip to main content

folkfox

Skip to main content
Skip to content
MICROSOFT ADS ENCYCLOPAEDIA

Predictive targeting and audience expansion

Predictive targeting uses machine learning to identify high-converting audience segments beyond your existing keyword and audience parameters.

Quick answerPredictive targeting in Microsoft Advertising uses machine learning to find new prospects with conversion characteristics similar to your existing converting customers. It analyzes landing page content, ad copy, and audience assets to discover incremental conversion opportunities.
Section 01

How Predictive Targeting discovers new audiences#

Predictive targeting uses deep neural networks to discover prospective customers who share behavioral, demographic, and search patterns with your proven conversion base.

As documented in Microsoft Predictive Solutions, 2026, the system analyzes converting search queries and landing page semantics to find relevant audience pockets outside manual targeting parameters.

Section 02

Audience expansion levels and controls#

Audience expansion operates on a dynamic slider, allowing advertisers to choose between conservative expansion (tight lookalikes matching core converting signals) and broad reach as explained in Microsoft Audience Expansion Docs, 2026.

Efficiency control

Pair predictive targeting with Target CPA bidding to ensure expanded impressions only serve when conversion probability meets cost goals.

Section 03

Lookalike audience modeling on first-party data#

When seed Customer Match lists or high-volume remarketing pools are provided, predictive algorithms build rich lookalike profiles, identifying matching consumers across the Microsoft Audience Network.

Section 04

When to enable or disable predictive expansion#

Enable predictive targeting on mature campaigns seeking incremental volume scale. Disable expansion on strict brand defense campaigns or niche B2B search ad groups requiring 100% deterministic keyword matching.

Consult our Microsoft Ads Hub for scaling frameworks.

Section 05

Predictive targeting deployment framework#

The table below summarizes optimal deployment conditions for predictive audience expansion.

Optimal deployment scenarios for predictive targeting
Campaign TypeExpansion SettingRecommended Guardrail
Audience Network NativeEnabled (Balanced)Strict Target CPA bid ceiling
Scale E-Commerce ShoppingEnabled (Conservative)Target ROAS threshold
Exact Match B2B SearchDisabled100% deterministic keyword match
Remarketing RecoveryEnabled (Lookalike Seed)Exclude past 30-day purchasers
Questions

Frequently asked questions#

What is Predictive Targeting on Microsoft Ads?

Predictive Targeting uses AI to automatically find new audiences who are likely to convert based on your current converting customer data and landing page content.

Can I disable Predictive Targeting on Microsoft Ads?

Yes, you can enable or disable predictive targeting at the ad group level under audience targeting settings.

Does Predictive Targeting increase my costs?

When paired with automated bidding strategies like Target CPA, predictive targeting only bids on auctions predicted to meet your efficiency targets.

What data does Microsoft use to build predictive audiences?

Microsoft analyzes your UET conversion history, landing page content, ad creative text, and first-party seed audience lists.

Is predictive targeting available on both Search and Audience networks?

Predictive targeting is primarily deployed across the Microsoft Audience Network and Performance Max, with algorithmic expansion signals informing search matching.

Keep reading

Read more on this topic#

Ready to scale incremental audience reach with AI?

folkfox configures predictive targeting and lookalike audience pipelines on Microsoft Advertising.