Systematic Audience Expansion for Horizontal Scaling

Hello!

In our last lesson, we built a troubleshooting framework to diagnose and fix efficiency drops that happen when you increase your budget. One of the key solutions we identified for audience saturation was to expand your audience. Today, we'll dive deep into the "how" of this solution by mastering a core scaling technique: horizontal scaling.

This lesson is all about systematically finding new, untapped groups of customers for your best-performing ads. Your goal is to expand audiences systematically for horizontal scaling, using broader lookalikes and new interests. We will cover practical, repeatable frameworks for expanding Lookalike audiences, discovering profitable new interest groups, and knowing when to let the algorithm take the lead with broad targeting. This is a critical component of the reusable scaling playbook you're building.

The Foundation: Horizontal Audience Scaling

First, let's solidify our understanding of horizontal scaling. While vertical scaling means giving more budget to a proven ad set, horizontal scaling involves taking a proven creative to a new audience. It's your primary tool for combating audience fatigue and unlocking long-term growth.

Vertical vs Horizontal Scaling At a Glance

This table provides a concise comparison of the two scaling methods.

Review the table under the heading "Vertical vs Horizontal Scaling At a Glance." Focus on the "Primary Goal," "When to Use It," and "Core Tactic" for horizontal scaling. This will set the context for the specific tactics we'll cover next.

The fundamental process for horizontal scaling is simple:

Identify a winning ad set (proven creative + audience).
Duplicate the ad set.
Change only one variable: the audience.
Launch and monitor.

Now, let's explore the three primary ways to change that audience variable: expanding Lookalikes, testing new Interests, and going Broad.

1. Systematic Lookalike (LAL) Expansion

Lookalike audiences are often the first stop for horizontal scaling. When your 1% LAL is performing well but you see frequency climbing, it's a clear signal to expand.

Meta Ads Lookalike Audience Creation Interface — The Lookalike Audience creation tool in Meta Ads Manager. Expanding your audience is as simple as moving the slider from 1% to a higher percentage like 3% or 5% to create a larger, broader audience.

However, how you expand is crucial for maintaining performance. A common mistake is to create a new ad set targeting a segmented 1-2% LAL, excluding the original 0-1%. This forces the new ad set into the learning phase from scratch. There's a better way.

How To SCALE Facebook Ads Lookalike Audiences

This video from Ben Heath demonstrates the most effective way to scale a Lookalike audience.

Watch these two key segments: First, understand the strategy of increasing the percentage (e.g., from 1% to 3%) rather than creating segmented audiences. Next, see how to apply this by editing the existing ad set. The key takeaway is that you edit your original, well-performing ad set and simply swap the 1% LAL for the new, broader 3% LAL. This preserves the ad set's social proof and optimization history.

A Tiered Lookalike Strategy

To make your LAL expansion more systematic, you can apply a tiered approach based on the quality of your source data. This gives you a logical roadmap for testing.

Implementing a Tiered Lookalike Strategy

This article from Silver Spoon Agency outlines a clear, tiered framework for building Lookalikes.

Read the section titled "Implementing a Tiered Lookalike Strategy." Internalize the concept of starting with your highest-intent audiences and expanding outwards.

Here's how you can structure your tiers:

Tier 1 (Highest Intent):
- Value-Based LALs: Built from a customer list that includes Lifetime Value (LTV). This tells Meta to find people similar to your best customers.
- "Super" Lookalikes: Built from a custom audience that stacks multiple high-intent signals (e.g., repeat purchasers + top 25% time on site).
- Purchasers LAL: A standard Lookalike from all customers.
Tier 2 (Medium Intent):
- Initiate Checkout LALs: People who have started the checkout process.
- Add to Cart LALs: People who have added a product to their cart.
Tier 3 (Lowest Intent):
- Top 25% Website Visitors LALs: Based on time spent on site.
- 95% Video Viewers LALs: People who watched nearly all of a key video ad.

Start by scaling your Tier 1 audiences. Once they are saturated, you can systematically move to testing Tier 2, and so on, always using your top-performing creatives.

2. Strategic Interest Expansion

Testing new interests can feel like a guessing game, but a strategic approach can turn it into a powerful tool for training the Meta algorithm. The goal isn't just to find new pockets of customers, but to broaden the algorithm's understanding of who your ideal customer is, which in turn makes your broad targeting campaigns more effective.

The secret is to think "one step removed."

The Best Facebook Ads Targeting Strategy for 2026

In "The Best Facebook Ads Targeting Strategy for 2026," Sam Piliero explains a powerful method for interest expansion.

Watch these three sections to understand the full strategy: First, grasp the core concept of combining broad and interest targeting to train the pixel. Next, focus on the "one step removed" targeting logic, explained with the Nike example from 06:02 to 09:13. Instead of targeting direct competitors (e.g., Adidas), you target ancillary interests (e.g., Range Rover) that your affluent customer might also have. Finally, understand which creatives to use in these new interest ad sets from 10:00 to 10:52. You use your best-performing ads to give these tests the highest chance of success.

Practical Application:
Imagine you're marketing a high-end, organic skincare brand.

Obvious Interests (already covered by Broad): "Skincare," "Sephora," "Beauty."
"One Step Removed" Interests to test: "Whole Foods Market," "Yoga," "Gwyneth Paltrow," "Organic food."

By getting conversions from these tangential interest groups, you teach Meta's algorithm that people interested in organic food and wellness are also great customers for your brand, thereby expanding the pool of people it will target in your main broad campaigns.

3. Graduating to Broad Targeting

The final frontier of audience expansion is to remove targeting constraints almost entirely. This is known as Broad Targeting (or "Advantage+ audiences" in Meta's terminology). It might seem counterintuitive, but at a certain point, it becomes the most effective way to scale.

Advanced Audience Expansion and Broad Targeting

This guide provides the criteria for when and why to use broad targeting.

Read the section "Advanced Audience Expansion," focusing specifically on the subsection "Broad Targeting for Scale." Then, review the FAQ entry for "When should I try broad targeting?"

Here are the key conditions for testing broad targeting:

You have strong conversion data: Your Meta Pixel should have at least 100-200 purchase events per month, giving the algorithm a clear signal of who to look for.
You have proven creative: You need ads that are known to convert well. With no audience targeting, the creative and the pixel data do all the work.
Your product has wide appeal: Broad targeting is less effective for extremely niche products.
You have exhausted initial scaling: You've already scaled your best LALs and interest groups.

The interest expansion work you did in the previous step directly fuels the success of broad targeting. By teaching the algorithm about new customer profiles, you make it smarter and more effective when you finally give it the freedom to operate in a broad environment.

Conclusion

You now have a systematic playbook for horizontal audience expansion, a crucial skill for achieving profitable scale. Instead of randomly testing audiences, you have a structured process.

Here are your key takeaways:

Horizontal scaling is about finding new audiences for winning ads to combat saturation and unlock growth.
Expand Lookalikes systematically. Edit your performing ad set to use a broader percentage (e.g., 1% -> 3%) to retain learning, and use a tiered approach based on source-data quality.
Expand interests strategically. Test "one-step-removed" interest groups to train the algorithm and improve the performance of your entire account.
Graduate to Broad Targeting when you have strong data. Once your pixel is well-seasoned and you have proven creatives, letting the algorithm take control is often the best way to scale further.

In our next lesson, we will focus on the structural side of scaling. Now that you know how to find new audiences and have winning ads, we will learn how to repurpose winning ad sets and creatives into new campaign structures to expand reach effectively.