Solved: ROAS Drops When Scaling Facebook CBO Campaigns

Hi there,

Thanks for reaching out.

I've had a good look through the situation you described with your ad campaigns. It's a classic problem and one I see all the time. You've found something that works on a small scale, but the moment you try to pour more fuel on the fire, the whole thing sputters out. It's incredibly frustrating, but the good news is that it's almost always solvable. The issue isn't really a simple "CBO vs. ABO" debate; it's a bit deeper than that. It's about understanding what you're actually telling the Meta algorithm to do when you press that 'increase budget' button and having a proper structure that's built for scale from the ground up, not just for pottering along at £50 a day.

I'm happy to give you some initial thoughts and guidance on this. The solution lies in shifting your mindset from just duplicating what works to building a repeatable system for testing, scaling, and knowing your numbers so well that you can confidently spend more because you know exactly what a customer is worth to you. It's less about finding a magic campaign setting and more about building a solid advertising engine for your business.

We'll need to look at the CBO vs. ABO myth...

Right, let's get this out of the way first because it seems to be the main thing on your mind. There's so much noise online about whether CBO (Campaign Budget Optimisation) or ABO (Ad Set Budget Optimisation) is 'better'. The truth is, one isn't better than the other; they are just tools for different jobs. Thinking one is inherently superior is a common mistake that costs advertisers a lot of money. It’s like asking if a hammer is better than a screwdriver. It depends entirely on what you're trying to build.

ABO is your screwdriver. It's for precision work. When you're testing new creatives, new audiences, or new ad copy, you need control. You want to give each ad set a specific, equal budget to ensure it gets a fair chance to perform. If you put three new audiences in a CBO campaign, the algorithm will almost immediately pick one it *thinks* will win (often the one with the cheapest clicks, not necessarily the best buyers) and pour most of the budget into it, leaving the others to starve. You'll never know if audience #2 or #3 could have been a goldmine with a bit more spend. ABO lets you force the budget where you want it, making it perfect for the testing phase. You gather clean, reliable data on which creative and which audience actually works, without the algorithm making assumptions for you.

CBO is your hammer. It's for when you've done the precision work and you just need to drive the nail home with force and efficiency. Once you have identified your winning creatives and winning audiences from your ABO tests, you move them into a CBO campaign. Here, you're telling Meta, "I've given you 3-5 proven ad sets. I trust you to find the cheapest and most effective conversions among them. Go and spend my £100 a day as efficiently as possible." The algorithm is brilliant at this. It will dynamically shift the budget in real-time to the ad set that's getting the best results at that moment, lowering your overall cost per purchase. Using CBO for untested assets is a recipe for wasted spend. Using it for proven assets is the key to efficient scaling.

Your experience confirms this. You saw worse performance with ABOs because you were likely using them in the wrong context, perhaps expecting them to perform like a scaling campaign. And your CBOs work well until you try to scale, because they're good at optimising a small budget but expose the weaknesses in your strategy when you ask them to do more heavy lifting.

To make this really clear, here's a simple flowchart that maps out the thinking process. This isn't a rigid rule, but it's the logic that underpins a stable advertising structure.

So, the first big change is to stop thinking about which one is 'better' and start building a system where you use both for their intended purpose. You'll have an 'always on' ABO campaign for testing, and a separate CBO campaign where only the champions from your tests get to compete for the real scaling budget.

I'd say you need to understand why scaling is really failing...

Your problem – "every time I tried higher budgets, performance dropped significantly" – is one of the most common complaints I hear. And the reason it happens has very little to do with the CBO/ABO setting. It has everything to do with what happens to your audience and creative when you increase the spend. You're not just turning up a volume dial; you're fundamentally changing the instructions you give the algorithm.

At £50/day with broad targeting, Meta is cherry-picking the absolute lowest-hanging fruit for you. It's finding the tiny fraction of the millions of people in that "broad" audience who are most likely to buy your product *right now*. They might have just visited a competitor's site, their birthday is next week, or they've recently engaged with similar products. They are red-hot buyers, and it doesn't take much to convert them. Your ROAS is high because the sales are easy.

The moment you increase the budget to £100, and especially when you do it suddenly by 20% or more, a few things happen:

1. You Exit the 'Easy Buyer' Pool: There are only so many of those low-hanging fruit customers available on any given day. To spend the extra budget, the algorithm has to push your ads out to a wider, colder audience. These people are less problem-aware, less ready to buy, and require much more persuasion. Your creative that worked on the hot audience might be completely ineffective on this new, more sceptical group.
2. You Trigger the Learning Phase: A significant budget edit (and yes, 20% is often significant enough) tells the algorithm that the conditions have changed, and it needs to re-evaluate its strategy. This often throws the ad set back into the 'Learning Phase'. During this time, performance is erratic and unpredictable by design. The algorithm is experimenting, showing ads to different types of people to figure out how to spend the new, higher budget effectively. Your performance tanks because the system is exploring, not exploiting. Constantly duplicating campaigns or making drastic budget changes means your campaigns are *always* in this unstable learning state and never get a chance to optimise properly.
3. Audience Saturation and Ad Fatigue: You're hitting the same small pool of people more often. Your ad 'frequency' metric likely shoots up. People see the same ad again and again, get bored of it, and stop paying attention. This is creative fatigue. The ad that was a winner for 3 weeks is now just background noise. Duplicating the campaign with the same creative doesn't solve this; it just puts the same tired ad in front of the same tired audience under a new campaign name.

The drop-off isn't a gradual slope; it's often a cliff. You go from profitable to losing money overnight, exactly as you described. It looks something like this:

So, the answer to "when is the best time to scale a CBO?" is not after 2-3 days or a certain amount of spend. The answer is: you scale when you have a winning ad set that has proven itself stable, is out of the learning phase, AND you are prepared to accept a lower ROAS. Or, more importantly, when you have a constant stream of new, tested creatives and audiences to feed the machine so you're not just hammering the same nail over and over again.

You probably should build a real scaling machine...

The solution isn't to stop spending more. It's to build a proper structure that anticipates these challenges. You need to separate your activities into distinct campaigns with distinct goals. This is how professional media buyers do it, and it's how you get stability.

Here’s a basic structure I would recomend:

With this structure, you're not guessing anymore. Your Testing Campaign is constantly feeding your Scaling Campaign with fresh, proven ad sets. When a winner in the Scaling Campaign starts to fatigue and its performance drops, you simply turn it off and replace it with the next graduate from your Testing Campaign. This creates a cycle of continuous improvement, not a cycle of panic-duplicating. Your scaling is done by adding more proven ad sets to the CBO or by slowly, gradually increasing the CBO budget (no more than 15-20% every 48-72 hours) once it's stable.

You'll need to know the only number that truly matters for scaling...

Here is the most important, and most overlooked, piece of the puzzle. The real reason advertisers fail to scale isn't because of CBO settings or creative fatigue. It's because they have no idea how much they can actually afford to spend to acquire a customer. They're obsessed with keeping ROAS as high as possible, which ironically keeps their business small.

You said your ROAS was 2.5-4. That's great. But when you tried to scale, it dropped to 1.0. That's obviously not sustainable. The question you need to answer isn't "How do I get my ROAS back to 4.0 at a higher spend?". The question is: "What is the lowest ROAS I can accept and still be profitable?". The answer lies in your Customer Lifetime Value (LTV).

LTV tells you the total profit you can expect to make from a single customer over the entire duration of their relationship with your business. Once you know this number, everything changes. You're no longer just trying to make a profit on the first sale; you're acquiring an asset that will pay you back over time. One campaign we worked on for a subscription box client achieved a 1000% return on ad spend. This wasn't just on the first box; it was possible because they understood the long-term value of a subscriber, which allowed us to bid more aggressively than their competitors.

Let's do the maths. It’s simpler than you think.

• Average Revenue Per Account (ARPA): What's the average amount a customer spends with you? If they only buy once, this is your Average Order Value (AOV). Let's say it's £80.
• Gross Margin %: What's your profit margin after cost of goods? Let's say it's 60%. So your gross profit per order is £80 * 0.60 = £48.
• Repeat Purchase Rate: How many times does an average customer buy from you in their lifetime? Let's say it's 1.5 times.

LTV = (AOV * Gross Margin %) * Repeat Purchase Rate

LTV = (£80 * 0.60) * 1.5 = £72

In this simplified example, each customer you acquire is worth £72 in gross profit to your business. This is your North Star. A common rule of thumb is to aim for a 3:1 LTV to Customer Acquisition Cost (CAC) ratio. This means you can afford to spend up to £24 (£72 / 3) to acquire a single customer.

Your CAC is your total ad spend divided by the number of new customers. If you're spending £120 and your ROAS is 1.3, you're making £156 in revenue. If your AOV is £80, that means you got roughly 2 customers (£156 / £80). Your CAC is £60 per customer (£120 / 2). This is way higher than your affordable £24 CAC. No wonder it's not working.

But what if you could improve your repeat purchase rate to 2.5 through email marketing? Your LTV jumps to £120. Now your affordable CAC is £40. Suddenly, you have more breathing room. What if you could increase your AOV to £100? Your LTV becomes £150, and your affordable CAC is £50. Now that £60 CAC doesn't look so terrifying. You're close to profitability even at that lower ROAS. This is how businesses scale. They don't just find cheaper ads; they increase the value of their customers.

To help you with this, I've built a simple calculator. Play around with your own numbers and see how small changes can radically alter how much you can afford to spend on ads.

My main recommendations for you are below:

Alright, that was a lot to take in. The key takeaway is to move from a reactive approach (duplicating campaigns when they break) to a proactive, systematic one. You need to become the architect of an advertising system, not just an operator pressing buttons. It requires more work upfront, but it's the only way to build something stable and truly scalable. We've used these principles to help eCom stores achieve incredible results, like a 691% return for a women's apparel brand and an 8x return for a maps & navigation company. It works because it's based on sound marketing principles, not chasing algorithmic trends.

Here is a summary of the main advice I have for you, laid out as an actionable plan.

This might seem like a lot of work, and frankly, it is. Managing a system like this, from creative production to data analysis and strategic budget allocation, is a full-time job. It's the difference between 'running ads' and professional media buying. It requires discipline and a deep understanding of the platform and your business metrics.

The fluctuations you're seeing aren't random; they're the logical outcome of your current approach. By adopting a more structured, strategic framework, you can turn that volatility into predictable growth. It takes the emotion out of it and turns advertising into what it should be: a measurable, scalable engine for your business.

If you're feeling a bit overwhelmed by all this, that's perfectly normal. This is complex stuff, and implementing it correctly while also running a business is a huge challenge. This is often the point where founders realise they need an expert partner to manage this for them, so they can focus on their products and customers.

We specialise in building and managing these kinds of scalable advertising systems for businesses just like yours. If you'd like to have a chat, we offer a free, no-obligation initial consultation where we can look at your ad account together and map out a more detailed, customised strategy for you. It might be the most valuable 30 minutes you spend on your marketing this year.

Hope this helps!

Regards,

Team @ Lukas Holschuh