The 6-Step Guide to Nailing & Scaling Creative Testing in Paid Social

If you’re running paid social campaigns and your performance is plateauing, chances are, it’s not your targeting. It’s not even your budget. It’s likely your creative.

Think about it. 

Everyone has access to the same audience tools, bidding strategies, and platforms. But creative? That’s where brands separate themselves from the competition. Meta states that creative accounts for 56% of a campaign’s success.

If your ads aren’t stopping thumbs, engaging users, and driving action, no amount of audience tweaking or budget shifting will save them. 

The brands that scale the fastest aren’t just the ones spending the most; they’re the ones testing, refining, and doubling down on creative that works.

Take a quick look at the Facebook Ad Library of any big-spending brand (like Bellroy below), and you’ll find hundreds of active ads running at once.

Testing at that scale might seem overwhelming, but the right creative testing framework can dramatically increase your output.

Before diving into testing techniques, let’s first break down best-practice campaign structures to ensure your tests are set up for success.

Step 1 – Setup The Right Account Structure for Creative Testing

When it comes to testing, how you set up your campaigns is critical, otherwise, your tests may end up being a huge waste of time. Most of the time when auditing accounts, we see brands make the mistake of dumping test ads into their main campaigns and letting the platform “figure it out.” 

Bad move. That’s a fast track to Meta Ads (previously Facebook Ads) not delivering your new ads and never getting feedback fast enough. 

Instead, follow this structure:

Facebook Ads creative testing structure

Testing Campaign: ABO (Ad Set Budget Optimisation)

When you’re testing creative in Meta Ads or other social platforms, you need clean data. That means setting up an ABO testing campaign(s) where each creative variation gets equal spend.

Why? Because Meta’s algorithm is designed to find a winner quickly, often favoring one ad over others before a fair test has run. An ABO campaign forces the platform to spend evenly, giving you a true read on which creative is working.

It’s best to test new creatives in audiences you have already validated to ensure you’re only testing one variable at a time. If you’re new to Meta Ads, then you’ve got a bit of work to do to find both winning audiences and creative. These days, broader audiences perform best so long as you’ve got a seasoned pixel. 

Your testing campaign should hold no more than ~20-30% of your total channel budget to ensure the majority of your ad spend is pushed behind the winners. Any more than that, and you’re risking turbulent performance. 

Scaling: CBO & Advantage+

Once you’ve found a winning creative, it’s time to put it into your scaling campaigns using Campaign Budget Optimisation (CBO) and/or Advantage+ Shopping campaigns. 

These campaigns hold at least ~70% of your budget, including your top-performing audiences and creative that you’ve previously validated. 

The idea is that it doesn’t matter where the platform weights spend, as you’ve previously validated the creatives.

Step 2 – Always Start With a Hypothesis

Before you run any test, ask yourself: What are we trying to learn?

A strong hypothesis gives structure to your testing. Instead of just launching random variations, define a clear goal:

• “Will using a question in the first 3 seconds increase thumbstop rates?”
• “Does a UGC-style video perform better than a product-only animation?”
• “Does a darker color scheme drive more engagement for luxury products?”
• “Do carousels outperform video at retargeting level” 

If you’re testing without a hypothesis, you’re just spending money without getting smarter. 

Step 3 – Gather Inspiration In/Out of Category

Are you currently opening up a blank document and trying to come up with new ad ideas? 

Just please stop. 

Before you jump into creating new ads, take time to gather inspiration from top-performing ads both inside and outside your industry.

We use Foreplay to save and analyse high-performing ad creatives across different industries. This allows us to:

• Identify trends in hooks, messaging, and formats that are driving engagement.
• Spot creative gaps in our own strategy by comparing to leading brands.
• Avoid reinventing the wheel by iterating on proven concepts.

To determine which ads are performing best, we look at brands that are spending big, like I said, the one running hundreds of active ads in Meta’s Ad Library. But more importantly, we focus on which ads have been running the longest. If an ad is still active after weeks or months, it’s likely performing well, otherwise the brand would have paused it.

Beyond individual ad performance, analysing these long-running ads helps spot trends across industries. What formats, headlines, and creative styles keep showing up? What’s working across different markets? This kind of insight helps shape our own testing roadmap.

The best ideas don’t always come from your industry. Looking at ads from different verticals can spark fresh concepts you might not have considered that will help you stand out.

Step 4: Run Tests for the Right Duration & Spend Level

Testing isn’t a one-day sprint. If you kill an ad too soon, you might be shutting down a winner before it has a chance to gain traction. On the flip side, running a poor ad for too long can drain your budget.

A good rule of thumb: Run tests for at least 3-5 days or at least 3x your target CPA (so if you needed sales for $50, you’d allow the ad to spend $150 before determining to pause or scale) to allow Meta’s learning phase to do its job.

Step 5: Benchmark Your Creative Performance

You can’t improve what you don’t measure. Before deciding whether a creative is a winner or a loser, compare it against your baseline metrics for the part of the funnel you are reviewing. 

What you expect from the prospecting part of your funnel is different from retargeting, so knowing your key benchmarks is essential. We recommend leveraging a reporting platform like Motion as it simplifies the analysis process at scale.  

 Some key metrics:

• CTR (Click-Through Rate): Aim for 1.2%+
• Thumbstop Ratio (3-Second Video View Rate): Should be 25%+
• CPM (Cost Per 1,000 Impressions): Lower CPMs against your baseline for that audience generally indicate higher engagement
• ROAS (Return on Ad Spend): Should align with overall account goals but keeping in mind that platforms over-attribute. Always review business level metrics like marketing efficiency ratio before deciding to scale. 

If an ad falls below these benchmarks, iterate. If it meets or exceeds them for an extended period, it’s time to do the slip and slide to your scaling campaigns.

Remember, though, that 90% of ads will likely be duds. So make sure you give yourself a high-five when popping that bad boy into your scaling campaign. 

That said, this isn’t a black-and-white rule. When reviewing new ad performance, also measure how your overall account is performing. If performance is under baseline across the account and your new ad is also under baseline, don’t kill it immediately. Consider other variables impacting results like seasonality, promo periods, and even broader market trends that can affect CPMs and conversions.

Understanding external factors ensures you’re making the right optimisation decisions rather than prematurely pausing ads that still have potential.

Step 6: Ship 70% Iterative Creative & 30% Net New

A big part of scaling creative effectively is balancing iteration vs. new concepts. We use a 70/30 approach:

• 70% of creative production focuses on iterations, refining and improving what’s already working.
• 30% is dedicated to net-new concepts, completely fresh ideas to test new angles.

Again, you do not want to fall in the trap of trying to pump out new ads all of the time. The best way to ensure a consistently high performing ad account is following this formula. 

Examples of Iterative Creative Testing:

• Breaking a high-performing video into a carousel format to see if engagement improves.
• Testing a new hook that calls out a different audience segment.
• Rearranging content blocks within a video and analysing how it impacts video completion rates.
• Tweaking the copy to try emojis vs. no emojis. 
• Adapting a static image into a motion-based ad to see if movement increases attention.

Iterating on winners ensures you’re not constantly starting from scratch while still introducing enough fresh creative to avoid fatigue.

Repeat the Process and Keep Learning

Creative testing isn’t a one-time effort, it’s an ongoing cycle. Once you’ve shipped new creative and reviewed its performance, go back to Step 2 and start again.

• Refine your hypothesis based on what you’ve learned.
• Gather new inspiration to keep ideas fresh.
• Continue testing and iterating to maximise performance.

The brands that consistently win in paid social typically have one dedicated person driving creative strategy. When ownership is unclear, execution suffers, because if everyone is responsible, no one truly is.

Need help optimising your creative testing? Let’s chat.