Solved: Adding New Creatives to a Winning CBO Campaign

Hi there,

Thanks for reaching out!

Happy to give you some initial thoughts on this. It's a classic situation, you've got something that works and the fear of breaking it is very real. The question of how to introduce new creatives into a winning CBO is probably one of the most common, and most commonly messed up, parts of scaling Meta campaigns. Get it wrong and you can poison a campaign that's been printing money for a year. Get it right and you can breathe new life into it and scale further.

The short answer is you should almost never touch a winning campaign directly. But the real issue here is probably bigger than just how to add a new creative. A campaign running for a year, even a "winning" one, is likely suffering from some level of fatigue and could be much more efficient. Let's get into how you should handle this specific situation and then zoom out to a more robust, long-term structure that'll stop you from being in this precarious position again.

We'll need to look at why your 'winning' campaign is a ticking time bomb...

Alright, let's be brutally honest. Having a single campaign that's been your top performer for "the best part of a year" sounds great on the surface, but from an expert's perspective, it's a massive red flag. It tells me you're not testing enough, and you're one algorithm shift away from your entire acquisition pipeline collapsing. That campaign isn't a fortress; it's a house of cards.

When you launch a campaign on Meta, the algorithm enters what's called the "Learning Phase". It's exploring who in your audience is most likely to take the action you want (e.g., purchase, lead, etc.) based on the specific creatives you've provided. It chews through your budget to figure this out. Once it exits the learning phase, it has a pretty solid "map" of how to get you cheap results. It knows which creatives work for which segments of your audience within your ad sets.

When you make what Meta calls a "significant edit" – and yes, adding a whole new creative is significant – you risk throwing the entire campaign back into the learning phase. The algorithm's map gets wiped. It has to start from scratch. I've seen it happen countless times: a campaign doing a 5x ROAS gets a "quick little tweak" and the performance drops off a cliff, sometimes never recovering. The algorithm that was finely tuned to deliver your two specific ad sets with their original ads is now confused. It has to re-allocate the CBO budget, re-learn what works, and in that process, your costs will almost certainly go up before they (maybe) come back down.

Your caution is justified. Touching that campaign directly is like performing surgery on a running engine. You might get away with it, but you're more likely to cause some serious damage. So, the first and most important principle is: protect the winner at all costs. We don't touch it. We use it as a control group, a benchmark to beat. Everything we do next is based on this principle.

I'd say you need a structured way to test without breaking things...

So, we're agreed: we don't touch the original campaign. What do we do instead? You've got two main options, one is the safe, methodical approach, and the other is a bit more aggressive. The one you choose depends on your budget and risk tolerance.

Method 1: The Control Freak's Clone (Safest Option)

This is my default recommendation and the one you should probably start with. It’s all about isolating variables so you know for sure if your new creatives are actually better.

1. Duplicate the Campaign: Go to your Ads Manager, select your winning CBO campaign, and click "Duplicate." Create just one copy. This creates an exact clone of the campaign, including its ad sets and ads.
2. Isolate the New Creatives: In this newly created duplicate campaign, go into each ad set. Turn OFF the original, old ads. Now, create your new ads within these ad sets. You can do this by copying the actual ad into each ad set if you want to preserve engagement (post ID method), or just build them from scratch. The result is the same: you have a campaign identical to your winner, but the only difference is the creative.
3. Launch and Monitor: Launch the new (duplicate) campaign with a controlled budget. It doesn't have to be the same as the original, you could start with 20-30% of the original's budget. Now you have a perfect head-to-head test. Your original campaign is chugging along, safe and sound. Your new campaign is running with the exact same settings, but with the new creative.
4. Compare the Data: After a few days (enough time to exit the learning phase and gather meaningful data), compare the key metrics. Look at CTR (Click-Through Rate), CPC (Cost Per Click), and most importantly, CPA (Cost Per Acquisition) or ROAS (Return On Ad Spend). If the new campaign with the new creatives is clearly outperforming the original, you've got a new winner.
5. Scale the Winner: If the duplicate campaign is the clear winner, you can then begin to slowly scale up its budget while you slowly scale down the budget of the original campaign. This ensures a smooth transition without any sudden shocks to your account's performance.

This method is methodical, safe, and scientifically sound. Its only downside is that it can be a bit slower. But slow and steady often wins the race in paid ads.

Method 2: The CBO Hunger Games (Aggressive Option)

This method is for when you have a bit more budget to play with and you want to find the absolute best creative as quickly as possible. You're leveraging the power of CBO to force your creatives to compete against each other in a new enviroment.

1. Duplicate the Campaign (Again): Same first step. Duplicate the winning CBO.
2. Create a Creative Showdown: In the new campaign, keep your two original ad sets as they are, with the old creatives inside them. Now, DUPLICATE those two ad sets within the same campaign. In these two *new* ad sets, turn off the old ads and add your new creatives.
3. Let CBO Decide: You now have one CBO campaign with four ad sets: two with the old proven creatives, and two with the new challengers. You launch the campaign and let CBO do what it does best: allocate the budget to the ad sets (and therefore, the creatives) that are performing the best.

Within a few days, you'll see where Meta is pushing the money. If it's all flowing to the new ad sets, you have a definitive answer. This method is faster and can be more efficient as the algorithm is doing the heavy lifting for you. The risk is that it's a bit messier, and you're spending money on the old creatives in a new learning environment, which might not be the most efficient use of budget if they are already fatigued.

You probably should be more worried about creative fatigue...

Now, let's get back to that "running for a year" thing. Audience fatigue happens when the same people see your ads too many times and start to ignore them. Your 'Frequency' metric (the average number of times a user has seen your ad) will tell you this. For prospecting campaigns, once you start getting a frequency of 3-4+, you're often in the danger zone. For retargeting, it can be higher, but there's still a limit.

When frequency goes up, a few things happen:

• Your CTR goes down because people are bored of your ad.
• Because your CTR is lower, your CPC goes up. Meta charges you more to show an ad people aren't clicking on.
• Because your CPC is up, your CPA goes up. It costs more to get a conversion.

It’s a slow, silent killer of campaign performance. Your campaign might still be "profitable," but I'd bet my house that its profitability has been slowly eroding over the last six months. You're leaving money on the table. I remember one of our SaaS clients had a campaign they loved, running for 8 months. Their cost per acquisition was around £100, which was acceptable. We came in, diagnosed severe creative fatigue, and launched a new testing campaign using the methods above. Within six weeks, we had found new winning creatives and built a new evergreen campaign that brought their blended CPA down to just £7. That's not a small improvement; it fundamentally changed the economics of their business. The only reason we could do that was by recognising the old "winner" was actually a tired workhorse that needed retiring.

Use the calculator below to get a rough idea of how fatigued your current ads might be. It's not a perfect science, but it's a good gut check.

You'll need a proper, grown-up campaign structure...

This whole situation – being scared to touch your one good campaign – comes from a flawed account structure. You shouldn't have just one "winning" campaign. A healthy ad account is an ecosystem, not a monolith. The structure I build for all our clients, whether they're eCommerce, SaaS, or lead gen, is based on the marketing funnel.

It looks something like this:

• 1. Top of Funnel (ToFu) - Prospecting Campaigns: The only job of these campaigns is to find new customers. This is where all your testing happens. You run CBO campaigns targeting your cold audiences: lookalikes, detailed interests, broad targeting etc. This is your creative testing ground. You'd have multiple ad sets, each with a different audience, and inside them you're constantly cycling new creatives in and out (using the duplication methods we've discussed). You are constantly trying to beat your best performers.
• 2. Middle of Funnel (MoFu) - Engagement/Warm Retargeting Campaigns: This is a campaign that targets people who have shown some interest but haven't gone deep into your site. Think video viewers, social media page engagers, or all website visitors from the last 30 days (excluding purchasers and add-to-carts). The ads here are different – maybe they handle common objections, show testimonials, or explain a key feature.
• 3. Bottom of Funnel (BoFu) - Hot Retargeting Campaigns: This campaign is your closer. It targets people who are on the verge of converting: they've added to cart, initiated checkout, or visited a key pricing page in the last 7-14 days. The ads are very direct: "Forgot something?", "Complete your purchase", maybe with a small incentive like free shipping if that's part of your strategy.

Why does this structure solve your problem? Because your testing (ToFu) is completely seperate from your proven retargeting funnels (MoFu/BoFu). You can be as aggressive as you like in your ToFu campaigns, launching and killing new creative tests every week, and it will never, ever disrupt the performance of your BoFu campaign that's reliably converting your hot audiences.

Your current approach is like having one employee who has to do everything: find new customers, nurture leads, and close deals. It's inefficient. The funnel structure is like having a specialised team: a marketing department for prospecting, a sales development rep for nurturing, and an account executive for closing. Each is optimised for one specific job. That's how you build a scalable, resilient advertising machine that isn't dependent on one "lucky" campaign.

So, to wrap this all up, your initial question was simple, but the answer opens up a bigger conversation about how to manage your account professionally for long-term, stable growth. Don't just patch the problem; fix the underlying structural issue.

I've detailed my main recommendations for you below:

This might all seem like a bit of a faff compared to just clicking 'edit' and adding a new ad, but this is the difference between amateur ad management and professional media buying. One approach chases short-term results and is inherently fragile; the other builds a predictable, scalable system for growth. Running these tests, analysing the data, and restructuring an account takes time and expertise. Making a mistake can be costly, not just in wasted ad spend but in lost momentum.

This is exactly the kind of strategic overhaul we specialise in. We help businesses move from relying on one-off "lucky" campaigns to building robust acquisition engines. If you'd like to have a chat and walk through your ad account together on a screen share, we offer a completely free, no-obligation initial consultation. We can take a look at your setup and give you some more specific, actionable advice.

Hope this helps!

Regards,

Team @ Lukas Holschuh