All articles
Google Ads
Meta Ads

How to Identify Scalable Ad Sets in Meta & Google

When you’re running ads, one of the hardest decisions isn’t launching campaigns — it’s deciding which ones to feed more money when extra budget shows up. The knee-jerk move is to throw dollars at the highest ROAS campaigns, but not every “top performer” has room to grow. Some are already maxed out. Others look efficient only because they’re running at tiny spend.

This distinction matters most when the stakes are high — say you’ve just been handed an extra $15,000 and need to decide where to put it. Scale the wrong campaigns, and you’ll burn cash without incremental returns. Scale the right ones, and you’ll unlock new revenue without wrecking efficiency.

That’s the challenge: how do you tell the difference between a campaign that’s truly scalable versus one that’s tapped out?

The Challenge

You’ve got an extra $15,000.
The goal: allocate it across Meta and Google to maximize ROAS.
 But the question is: which ad sets actually deserve that budget?

Why This Challenge Matters

Scaling blindly is expensive. Without data, teams:

  • Overfund ad sets that look efficient but are already saturated
  • Miss high-ROAS campaigns flying under the radar due to low initial spend
  • Burn budget on "broad" campaigns that don’t return incrementally with scale

The cost? Wasted spend, missed revenue, and slower growth.

The Smart Way: What Prism Did

Prism took this prompt and returned a complete, performance-backed budget strategy—based on actual spend, revenue, creative fatigue, and audience responsiveness over the last 7 days.

It didn’t just identify high ROAS campaigns. It identified scalable ones.

How to Manually Identify Which Ad Sets Deserve More Budget

If you’re not using Prism, here’s the step-by-step method to run this analysis yourself:

1. Pull Your Last 7 Days of Campaign Data

From Meta Ads Manager and Google Ads:

  • Spend
  • Impressions
  • Clicks
  • Conversions
  • ROAS (preferably from a source like Appsflyer or GA4)

2. Flag High Performers

Look for ad sets with:

  • ROAS ≥ 2.5
  • CTR improving or stable
  • Frequency < 3 (for Meta)
  • Impression Share < 80% (for Google)
  • Consistent conversions (no flukes)

3. Run a Scalability Check

Ask:

  • Has ROAS held steady as spend increased?
  • Are we still reaching new audiences?
  • Is the creative fatigued (declining CTR, higher CPA)?
  • Are there lookalike or affinity audiences we haven’t tested yet?

4. Create a Tiered Reallocation Plan

Example:

  • Tier 1 (Scale Immediately)
    • Meta: Broad audience with UGC creative, ROAS 3.2
    • Google: Branded search with 70% impression share, ROAS 4.1
       
  • Tier 2 (Test Before Scaling)
    • Meta: Retargeting carousel with 1.8 ROAS but < 1 frequency
    • Google: Smart Shopping with improving ROAS but low volume

Distribute the $15K across Tier 1 first, then gradually feed Tier 2.

What to Avoid

  • Scaling campaigns with high ROAS but tiny volume → may not scale efficiently
  • Funding creatives that have peaked → check drop in CTR or increase in CPM
  • Prioritizing platform-reported metrics → always align with revenue, not just CPA

How Prism Decides Which Ad Sets Get More Budget

Prism doesn’t just glance at last week’s ROAS and make guesses.
 Here’s what it pulled together behind the scenes:

 Data Sources Used:

  • 30-day account performance overview
  • Campaign-level ROAS trends
  • Age group performance breakdown
  • Spend vs. return analysis by platform
  • Creative engagement patterns

With all this, it built a reallocation strategy — and suggested exactly which types of campaigns to scale next.

Prism’s Recommendations:

1. Lookalike Audiences
 → Built from recent purchasers; ideal for scaling cold reach with high intent

2. Retargeting Campaigns
 → Targeting cart abandoners or site visitors who didn’t convert

3. Broad Targeting with CBO (Campaign Budget Optimization)
 → Especially for high-performing age groups (e.g., 35+), where CPM is steady and CTR is rising

Each of these isn’t just a best practice — it was linked directly to recent performance insights inside the account. Prism caught what was working and pointed out where extra spend could actually stretch further.

 Why These Recommendations Work

These campaign types hit three key criteria:

  • Proven recent performance (backed by Appsflyer revenue data)
  • Available audience volume or remarketing pool
  • Creative assets that are still fresh (no fatigue)

So instead of spreading $15,000 thin across dozens of campaigns, you put it into three high-conviction bets — each backed by multi-layered analysis.