LinkedIn Sunset Lookalike Audiences: A Complete Guide to the New Era of Ad Targeting

On February 29, 2024, LinkedIn discontinued Lookalike Audiences. Learn how to use Predictive Audiences and Audience Expansion to maintain and improve your LinkedIn advertising performance.

On February 29, 2024, LinkedIn officially discontinued its Lookalike Audiences feature, marking a significant shift in how marketers can reach new audiences on the platform. This change affects every advertiser who relied on Lookalike Audiences to expand their reach and find prospects similar to their best customers. Understanding what replaced this functionality--and how to leverage it effectively--is now essential for any social media marketing strategy.

This guide explores the sunset of LinkedIn Lookalike Audiences, the two new targeting options that emerged in their place, and actionable strategies to maintain and improve your LinkedIn advertising performance.

What Was LinkedIn Lookalike Audiences?

LinkedIn Lookalike Audiences was a powerful targeting feature that allowed advertisers to reach new professionals who shared similar characteristics with their existing customers or website visitors. By analyzing the attributes of a source audience--such as job titles, industries, company sizes, or engagement patterns--LinkedIn's algorithm would identify and target LinkedIn members who exhibited similar professional profiles.

This feature became popular among B2B marketers because it offered a scalable way to expand reach beyond immediate networks while maintaining targeting relevance. Rather than manually building out complex audience segments based on dozens of individual attributes, Lookalike Audiences simplified the process by letting machine learning do the heavy lifting.

Key Capabilities of Lookalike Audiences

  • Automated audience building: Used AI to find similar professionals
  • Scalable targeting: Expanded reach beyond direct customer lists
  • Multi-attribute matching: Considered various professional characteristics
  • Simplified campaign setup: Reduced manual targeting complexity

The discontinuation of Lookalike Audiences reflects LinkedIn's broader investment in more sophisticated AI-driven targeting that focuses on behavioral prediction rather than static attribute matching. This evolution presents opportunities for advertisers who approach the new features strategically.

The LinkedIn Lookalike Audiences Sunset: What Happened and Why

On February 29, 2024, LinkedIn discontinued Lookalike Audiences as part of a broader evolution of its advertising platform. Existing Lookalike Audiences stopped refreshing and became static representations of the audience data at the time of sunset. Active ad sets continued to deliver using these static audiences, but no new members could be added based on similar characteristics.

According to LinkedIn's official announcement, the sunset aligned with the platform's investment in more sophisticated AI-driven targeting capabilities and privacy-conscious approaches to audience building.

Key Implications for Advertisers

For Active Campaigns: Ad sets using Lookalike Audiences continued running but with static audience data. This meant audiences would not grow or adapt over time, potentially reducing campaign effectiveness as market conditions changed.

For Future Targeting: New Lookalike Audiences could no longer be created, and existing ones could not be edited. Advertisers needed to rebuild targeting strategies using the new options LinkedIn introduced.

For API Integrations: The Lookalike Audiences API was also sunset, affecting marketing platforms and tools that relied on this functionality for automated audience management.

Important Timeline Dates

DateEvent
February 29, 2024Lookalike Audiences officially discontinued
March 30, 2024Inactive Lookalike Audiences archived
OngoingPredictive Audiences and Audience Expansion become recommended alternatives

The transition from Lookalike Audiences to Predictive Audiences represents LinkedIn's shift from static attribute matching to dynamic behavioral prediction. This evolution offers opportunities for more precise targeting when combined with quality first-party data.

Predictive Audiences: LinkedIn's New AI-Powered Solution

Predictive Audiences represents LinkedIn's next-generation approach to finding prospects likely to take action based on your existing customer data. Unlike Lookalike Audiences, which matched members based on static attributes, Predictive Audiences uses LinkedIn's AI to analyze behavioral patterns and identify professionals most likely to convert or engage with your ads.

According to LinkedIn's official documentation, this approach offers several advantages over the original Lookalike Audiences.

How Predictive Audiences Work

When you create a Predictive Audience, you select a source audience from your existing data--typically a list of customers, website visitors, or leads. LinkedIn's AI then examines this source group to understand not just who they are, but what actions they've taken and what behaviors indicate high intent. The algorithm then predicts which other LinkedIn members are most likely to exhibit similar behaviors.

This shift toward AI-powered targeting mirrors broader trends in AI automation for marketing, where machine learning enhances audience selection and campaign optimization.

Key Advantages Over Lookalike Audiences

Behavioral Focus: Predictive Audiences emphasize what members do rather than who they are, potentially identifying high-intent prospects that might be missed by traditional demographic or firmographic targeting.

Daily Refresh: Unlike the static nature of former Lookalike Audiences, Predictive Audiences refresh daily, ensuring targeting evolves with changes in member behavior and market dynamics.

Multi-Signal Analysis: The AI considers numerous signals simultaneously, including engagement patterns, content consumption, profile updates, and professional activities.

Requirements for Creating Predictive Audiences

Data Source TypeMinimum RequirementMaximum Limit
Contact/Company Lists300 rows300,000 rows
Lead Gen Forms300+ membersNo strict limit
Conversion Audiences300+ membersNo strict limit
Retargeting Audiences300+ membersNo strict limit

Important Limitations:

  • Maximum of 100 Predictive Audiences per ad account
  • Cannot be shared across ad accounts
  • Cannot be created from lists shared through Business Manager
  • Audience Expansion is disabled for ad sets using Predictive Audiences

Early results from advertisers show promising performance improvements. According to 10fold's analysis, some campaigns experienced meaningful improvements in click-through rates when switching from Lookalike to Predictive Audiences, though individual results vary based on industry, audience quality, and campaign objectives.

Early Performance Results

25%

Potential CTR increase reported by some advertisers

300+

Minimum members required for Predictive Audiences

100

Maximum Predictive Audiences per ad account

Best Practices for Predictive Audiences

Maximize your targeting effectiveness with these proven strategies

Source Data Quality Matters Most

The predictive model's accuracy depends entirely on the quality of your source data. Ensure your contact lists, website visitors, or conversion audiences represent your ideal customers.

Segment by Intent Level

Create separate predictive audiences for different segments. High-intent leads may produce different predictions than lower-funnel prospects.

Exclude Your Seed Audience When Appropriate

If you want to reach entirely new prospects, exclude your seed contact or company list from the targeting.

Monitor and Iterate

Since Predictive Audiences refresh daily, track performance over time. Consider refreshing source data or creating new audiences if results decline.

Audience Expansion: Scaling Your Existing Targeting

While Predictive Audiences focus on finding new prospects based on your data, Audience Expansion takes a different approach by automatically broadening your existing targeting criteria to reach similar professionals.

According to LinkedIn's official documentation, this feature is particularly valuable when you want to maintain your core audience definition while gaining additional scale.

How Audience Expansion Works

When you enable Audience Expansion for an ad set, LinkedIn analyzes your selected targeting criteria--such as job titles, skills, interests, or company attributes--and automatically expands to reach members who share similar professional characteristics. The expansion considers professional demographics as the primary data source.

When to Use Audience Expansion

Scaling Proven Campaigns: When you've identified targeting that works but need more reach to improve efficiency and reduce costs, Expansion provides scale without sacrificing relevance.

Awareness Campaigns: For campaigns focused on reach and brand awareness, Expansion offers an efficient way to maximize impressions.

Testing New Audiences: Use Expansion to reach a broader audience when testing new creative or messaging.

Important Limitations

  • Not available for dynamic ad formats or Predictive Audiences
  • Performance metrics include both core targeting and expansion audience
  • Excluded attributes from targeting won't be included in expansion
  • Audience count preview doesn't include expansion members

Audience Expansion works alongside your Matched Audiences--contact lists, website visitor audiences, and company followers--providing scale while retaining the specificity of your core targeting.

Predictive Audiences vs Audience Expansion Comparison
FeaturePredictive AudiencesAudience Expansion
Primary FocusBehavioral predictionAttribute-based scaling
Refresh RateDailyReal-time
Best ForConversion campaignsAwareness campaigns
Source Data RequiredYes (300+ members)No
Max per Account100 audiencesUnlimited
Works with ExpansionN/AYes
Available for Dynamic AdsNoNo

Migration Strategies: Moving from Lookalike to Predictive Audiences

If you were using Lookalike Audiences before the sunset, here's how to effectively transition your campaigns:

Step 1: Audit Your Existing Lookalike Audiences

Document your existing Lookalike Audiences and their performance metrics. Identify which ones delivered the best results and what source audiences you used. This information guides your transition strategy.

Step 2: Identify the Right Replacement Option

Choose Predictive Audiences when:

  • You have quality source data
  • Finding high-intent prospects is your priority
  • Campaigns focus on conversions or lead generation

Choose Audience Expansion when:

  • You want to scale existing targeting criteria
  • Campaigns prioritize reach and awareness
  • You prefer more control over core targeting

Step 3: Recreate Your Targeting

Create new Predictive Audiences or configure Audience Expansion based on your strategy. For Predictive Audiences, use your original source data where possible.

Step 4: Test and Optimize

Launch campaigns with new targeting and compare results to historical performance. Be prepared to iterate and refine your approach.

Common Migration Mistakes to Avoid

  1. Using Poor-Quality Source Data: Poor source data produces poor predictions
  2. Ignoring the 30-Day Archive Window: Archived audiences may be inaccessible
  3. Assuming Direct Equivalence: Predictive Audiences work differently and may produce different results

The transition underscores the importance of first-party data for modern advertising. As privacy regulations tighten and platform tracking capabilities evolve, first-party data--information customers directly provide--becomes increasingly valuable for targeting and personalization. Complement your LinkedIn strategy with a comprehensive SEO services approach to maximize organic and paid reach.

For deeper insights into LinkedIn's advertising tools, explore our guide on hidden Sales Navigator features that can enhance your B2B lead generation efforts.

Frequently Asked Questions

Ready to Optimize Your LinkedIn Advertising Strategy?

Our team of LinkedIn advertising experts can help you navigate the transition to Predictive Audiences and maximize your campaign performance.