Solved: Scaling New Winning Creatives on Facebook Ads

Hi there,

Thanks for reaching out! Happy to give you some of my initial thoughts and guidance on your question about scaling new winning creatives on Facebook. It's a really good question, and honestly, it's the exact point where a lot of advertisers either make or break their accounts. Getting this right is how you go from getting decent results to really significant growth.

You've got a solid base with a working CBO campaign, and you've done the right thing by having a seperate testing campaign. That already puts you ahead of many. Now, let's look at how to properly introduce those new winners into the mix without wrecking what's already working.

I'd say you need to weigh up the pros and cons of your two options...

You've basically laid out the two most common paths: either add the new creatives to your existing scaling CBO, or create a brand new CBO campaign with them. Both have their place, and the 'right' answer often depends on budget, how much risk you're willing to take, and how stable your current campaign is.

Option 1: Duplicating the new creatives into your existing scaling CBO campaign.

This is usually the quickest way to get data. The main advantage is that you're putting your new creatives into a 'hot' environment. The campaign is already optimised, the ad sets have exited the learning phase, and Facebook's algorithm has a very good idea of who to show your ads to. Your new creatives should start getting impressions and clicks pretty much straight away.

However, there's a definate down side. Any time you make a significant edit to a live ad set – and adding new creatives counts as significant – you risk disrupting its learning. It can sometimes push the ad set back into the learning phase. This isn't always a disaster, but it can cause performance to get a bit wobbly for a few days. You might see your CPA go up temporarily while the algorithm figures things out again. There's also a risk of what I call 'creative dilution'. If your four new ads are brilliant, but one of the old ones is still a superstar, Facebook might continue to pour the majority of the budget into that one old ad, and your new ones might not get a proper chance to shine.

I remember this happening with a client running a subscription box campaign. The campaign was delivering a solid 1000% Return On Ad Spend. It was very stable. We introduced a new set of UGC-style videos into the main CBO. While two of them eventually became top performers, the initial introduction caused a 20% dip in ROAS for about four days as the CBO re-allocated budget. It recovered, but it's a nervy few days when you're spending a decent amount.

Option 2: Creating a new CBO campaign with the same audiences but with the new creatives.

This is the safer, more scientific approach. You keep your existing, profitable campaign running as your 'banker'. It keeps bringing in the results, untouched and stable. Your new CBO campaign acts as a challenger. It's a straight shoot-out: old campaign vs. new campaign. This is a much cleaner test.

The main drawback here is that the new campaign has to start from scratch. It'll go through the full learning phase, which costs time and money. It might take a week or more to get optimised and for you to have enough data to confidently say whether it's actually better than your original campaign. If your budget is tight, running two large scaling campaigns in parallel might not be feasible. You're essentially doubling your spend for the duration of the test.

I remember working with a client selling online courses, we used this exact method. Their main CBO was generating sales. We built a challenger CBO with new video ads. For the first five days, the new campaign's ROAS was only about 200%. It looked like a failure. But after day seven, once it was out of the learning phase, it stabilised at over 447% ROAS in one week. We were then able to turn off the original campaign with full confidence. If we'd just dropped those ads into the old campaign, we might have killed them off after a few poor-performing days, thinking they were duds.

So, what's the verdict? If your budget is smaller and your current CBO isn't spending a massive amount, I'd probably lean towards carefully adding the new creatives to the existing ad sets. If you have a larger budget and protecting the perfomance of your current campaign is the top priority, then building a new challenger campaign is the more robust, professional method.

We'll need to look at your overall campaign structure...

Taking a step back, the fact you're asking this question suggests we could probably think about a more permanent, scalable structure for your account. Constantly testing in one campaign and then figuring out how to add winners to another can get messy. A better system is to structure your account based on the marketing funnel. This makes scaling much more systematic.

I usually build accounts out with campaigns dedicated to each part of the funnel. This is a simplified version, but the logic holds for most businesses, especially eCommerce.

-> ToFu (Top of Funnel - Prospecting): This is your cold traffic. The goal here is to find new customers. This is where you test your broad audiences, your interest-based audiences, and your lookalike audiences. Your scaling CBO with the five ad sets sounds like it's a ToFu campaign.

-> MoFu (Middle of Funnel - Consideration): These are people who have shown some interest but haven't taken a key action yet. This is your warm traffic. Audiences would include all website visitors (in the last 30-60 days), social media engagers, and people who watched a good chunk of your video ads.

-> BoFu (Bottom of Funnel - Conversion): This is your hot traffic, people who are very close to buying. Audiences here are your highest-intent users: people who have added products to their cart, initiated checkout, or viewed specific product pages in the last 7-14 days. You hit these people with strong call-to-actions, discount codes, or reminders about what they left in their cart.

When you structure your account like this, scaling becomes clearer. Your creative testing campaign is great. When you find a winning creative, you don't just ask "where does this go?". You ask, "what is this creative's job?". Is it a great ad for attracting brand new customers? If yes, it goes into your ToFu Prospecting CBO. Is it a great testimonial ad that would persuade someone on the fence? Maybe that goes into your MoFu Retargeting campaign.

This way, you have evergreen campaigns for each part of the funnel. Your job then becomes feeding these campaigns with your winning creatives and audiences, and turning off the ones that fatigue over time. It's a more sustainable system than the 'launch and duplicate' cycle. I recall a client in the outdoor equipment space, where implementing this structure allowed us to build a system that drove over 18,000 website visitors consistently every month. The prospecting campaign's job was to fill the funnel, and the retargeting campaigns' job was to convert them. Each had its own creatives tailored for that specific job.

You probably should consider how audiences affect scaling...

Scaling isn't just about finding more winning creatives. In fact, I'd argue that audience strategy is even more important once you have a set of proven ads. You can have the best ad in the world, but if you show it to the wrong people, it'll do nothing. Your current structure has five ad sets in one CBO, which is a good start for testing, but true scaling comes from expanding your audiences methodically.

The best audiences are almost always built from your own data. This is where Lookalike Audiences (LALs) come in. But not all LALs are created equal. I see so many accounts where people just create a 1% Lookalike of "All Website Visitors" and call it a day. You need to be more granular and prioritise based on intent. Think about the funnel we just discussed.

I would prioitise creating and testing lookalikes in this order:

1. Highest Value Customers: If you can, export a list of your top 20% of customers (by lifetime value) and create a lookalike from that. These are your VIPs, and finding more people like them is gold.
2. All Purchasers: This is the next best thing. A lookalike of everyone who has bought from you.
3. Initiated Checkouts: People who got right to the payment page. Very high intent.
4. Add to Carts: The next level down. Still a very strong signal of interest.
5. Website Visitors / Video Viewers: This is the broadest custom audience, so the lookalike will be the least potent, but it gives you massive scale.

You need at least 100 people in a source audience to make a lookalike, but honestly, you want more like 1,000+ for it to be really effective. As your pixel gathers more data, you should be systematically building and testing lookalikes from all these high-intent events.

Then, once your pixel is really mature (thousands of purchase events), you can start testing broad targeting. This is where you give Facebook an entire country, maybe layered with just age and gender, and let the algorithm do the work. It sounds crazy, but with enough data, it can be the most profitable way to scale because you're giving the algorithm maximum freedom.

You'll need a solid understanding of your numbers...

This is a point that gets missed. As you push to scale your ad spend, your costs will go up. Your Cost Per Acquisition (CPA) will rise. This is not a sign of failure; it's a normal economic reality. You first reach the 'cheapest' and easiest to convert customers. To find more, you have to reach further and spend more, so your average cost increases. The goal isn't to keep your CPA flat as you double or triple your budget – that's impossible. The goal is to manage the increase so that you remain profitable.

This is why understanding your business's numbers is so important. What is your average order value (AOV)? What is your customer lifetime value (LTV)? If your average CPA is £20, but your average LTV is £200, you have a huge amount of room to scale. Your CPA could climb to £40, £50, or even £60, and you'd still be running a very profitable business.

I remember a software client who hit a plateau. They couldn't spend more without their ROAS dipping below their target. The solution wasn't just in the ads. We worked with them on two things: improving their checkout flow to increase the conversion rate, and adding a higher-tier subscription plan to increase their LTV. By increasing the money they made per customer, they could suddenly afford to pay more to acquire a customer. This unlocked the next level of scale for their ad campaigns.

When you're looking at your scaling CBO, don't just judge it on the daily CPA. Look at the bigger picture. Are the customers you're acquiring sticking around? Are they buying again? Answering these questions is the key to building a truly sustainable and scalable advertising programme.

To bring all this together, scaling is a process of systematic testing and expansion, not just a single action. It involves your creatives, your audiences, your campaign structure, and your business economics. Below is a summary of how I might approach this if I were in your position.

This is the main advice I have for you:

As you can probably see, scaling effectively is a bit more involved than just dropping new ads into a campaign. It's about building a robust, repeatable system. While you can definately implement all of this yourself, this is where having an expert eye can really accelerate things. We've run these kinds of processes for all sorts of clients, from eCommerce stores selling apparel to B2B software companies chasing enterprise leads.

The value isn't in some secret trick, it's in the rigorous process and the experience of having seen these situations play out hundreds of times across different accounts. It helps you avoid costly mistakes, interpret the data correctly, and make the right decisions faster.

If you'd like to go through your account together on a call and map out a specific plan based on your exact data and goals, we offer a free initial consultation to do just that. It's a no-obligation way to get a second expert opinion and see if we might be a good fit to help you reach that next level of scale.

Hope this detailed breakdown has been helpful for you!

Regards,

Team @ Lukas Holschuh