Solved: CBO Campaign CPC Spike After New Ad Set Launch

Hi there,

Thanks for reaching out!

Happy to give you some initial thoughts on the CBO campaign issue you're seeing. It's a really common problem, and honestly, one of the biggest myths about Meta's platform is that CBO is some sort of magic button that stops you from bidding against yourself. The truth is a bit more complicated, and it nearly always comes down to how your audiences and campaigns are structured. Your CPCs shot up because you almost certainly started a bidding war with yourself, even inside a single campaign.

Let's get into why this happens and what you can actually do about it.

We'll need to look at why CBO isn't a magic wand...

Okay, so the first thing we need to clear up is what CBO (Campaign Budget Optimisation) actually does, versus what people *think* it does. Meta sells it as a tool that "distributes your budget to the top-performing ad sets in real time." And it does do that. If you have three ad sets and one is getting cheaper results, CBO will gradually shift more of the daily budget towards that winning ad set. So far, so good.

The problem is, it does nothing to solve the underlying issue of audience overlap. Imagine you have two ad sets. Ad Set A targets people interested in "Shopify". Ad Set B, your new broad one, targets... well, everyone in the UK aged 25-55. A huge number of people interested in "Shopify" are also, of course, in the UK and aged 25-55. So now you have two of your own ad sets trying to show an ad to the exact same person.

Meta's auction is just that: an auction. When two advertisers want to reach the same person, they bid against each other. When *you* are both of those advertisers, you're just artificially inflating the price you have to pay. You're telling the system, "I'm willing to pay X for this user in Ad Set A" and simultaneously saying, "I'm willing to pay Y for this same user in Ad Set B". The algorithm doesn't see "one advertiser"; it sees two distinct bids entering the auction for the same impression. The result? You pay more than you should have. CBO will still try to pick the "winner" between your two ad sets for that specific person, but the very act of them competing has already driven up the base cost for that impression. It’s a subtle but absolutely fundemental flaw in how most people set up their campaigns.

Think of it like sending two of your own salespeople to bid for the same contract at an auction house. They might both have the same company credit card, but by bidding against each other, they only succeed in making the final price higher for the company. That's what's happening inside your CBO campaign right now.

This is a simplified view of what's happening under the hood. The diagram below shows how adding that new, overlapping ad set forces your campaign into a state of internal competition, which is the direct cause of your rising CPCs.

I'd say you're probably bidding against yourself, and it's costing you dearly...

So, the core of the issue is audience overlap. The tricky part is that it's often invisible until you see the impact on your metrics. You might think "broad" targeting is completly seperate from a specific interest-based audience, but in reality, the overlap can be massive. Meta's algorithm is designed to find converters, and if it's found a pocket of high-intent users for your first ad set, you can bet it's going to try and target those same users with your new "broad" ad set because it already knows they are likely to perform well.

This is particularly true if your pixel is well-seasoned. The algorithm has a very good idea of who your customer is. When you give it a "broad" audience, it doesn't just randomly pick people. It uses the pixel data to find people within that broad demographic who *look like* your existing customers. And who looks most like your existing customers? The people you were already targeting successfully in your other ad set!

This is where things get expensive. Every time you add another ad set that isn't surgically excluded from your others, you're not expanding your reach as much as you are increasing the density of your bids on the same core group of people. I remember one client, a medical job matching SaaS platform, that came to us with this exact problem. Their campaigns were a complex web of overlapping audiences, driving their cost to acquire a new user up to around £100. By implementing a clean, funnel-based structure and eliminating that internal competition, we brought their cost per acquisition down to just £7. They were simply paying a huge premium to bombard the same pool of candidates from multiple angles.

The financial impact of this can be significant. A small percentage of audience overlap doesn't sound like much, but when you're dealing with an auction system, it can have an exponential effect on your costs. Use the calculator below to get a rough idea of how much this self-competition might be costing you. It's an illustrative tool, of course, but it demonstrates the principle. Even a 20-30% overlap can cause a significant jump in your effective CPC.

You probably should rethink your entire campaign structure...

The fact this happened tells me you're likely not structuring your campaigns with the user journey in mind. Just throwing different audiences into a CBO campaign and hoping the algorithm sorts it out is a recipe for wasted spend. A professional, scalable setup is built around the classic marketing funnel: Top of Funnel (ToFu), Middle of Funnel (MoFu), and Bottom of Funnel (BoFu).

Why is this so important? Because a user who has never heard of you (ToFu) needs a completely different message and has a different value to your business than someone who has already visited your pricing page and added a product to their cart (BoFu). By lumping them together, or allowing your targeting to overlap, you're treating them as the same, and the algorithm gets confused. It might optimise for cheap clicks from a cold audience while ignoring the high-intent user who is moments away from converting, simply because that user is more expensive to reach.

Here’s how I would structure it. This is the exact approach we use for our clients, from small e-commerce stores to large B2B SaaS companies. It creates clean data, prevents self-competition, and allows you to scale logically.

• ToFu (Top of Funnel - Prospecting): This is for finding new people. Your audiences here are broad, interest-based, and lookalikes of website visitors or video viewers. The goal is education and awareness. The key here is that you MUST exclude all of your existing website traffic and converters from this campaign. You are ONLY fishing for new blood.
• MoFu (Middle of Funnel - Retargeting): This is for people who have shown some interest but haven't taken a high-intent action. This includes people who have visited your website, watched a certain percentage of your videos, or engaged with your social profiles. Here, you're building trust and encouraging a deeper look. You'll want to exclude anyone who has already added to cart, initiated checkout, or purchased.
• BoFu (Bottom of Funnel - Re-engagement/Conversion): This is your highest-intent audience. These are the people who have visited the checkout page, added a product to their cart, or abandoned the payment process. The goal here is direct conversion. The messaging is urgent and focuses on overcoming final objections (e.g., offering a discount, reminding them of benefits). You ONLY target these specific high-intent users.

By separating these into different campaigns (or at the very least, different ad sets with rigorous exclusions), you create a machine. The ToFu campaign feeds new users into the MoFu campaign, which then feeds qualified prospects into the BoFu campaign. Each stage has a clear objective, and the audiences never overlap. This means no more bidding against yourself. CBO can then work as intended *within each campaign*, optimising budget across your different ToFu audiences, for example, without them clashing with your BoFu retargeting efforts.

This disciplined approach is what unlocks significant returns. For instance, for one of our clients, a women's apparel brand, we managed their Meta and Pinterest campaigns to achieve a 691% return on their ad spend. Results like these don't come from letting the algorithm guess; they come from building a clean, logical structure that prevents wasted spend and ensures the right message reaches the right person at the right time.

Below is a visual representation of how this funnel structure should look inside your ad account. Notice the flow of users and, most importantly, the exclusion rules at each stage.

You'll need a clear and ruthless exclusion strategy...

Let's get practical. The diagram above is the strategy, but the implementation happens inside Ads Manager. An exclusion strategy isn't a "nice to have"; it is the single most important technical step to prevent the issue you're facing. You need to be ruthless about it.

Every single prospecting (ToFu) ad set you ever create must, without exception, have your retargeting audiences excluded. This means creating custom audiences for "All Website Visitors (180 Days)", "All Customers/Purchasers", and "All Facebook/Instagram Page Engagers (365 Days)" and putting them in the exclusion box of your ToFu ad set.

This simple action tells Meta: "Go find me new people, but under no circumstances are you to spend my money on anyone who already knows who I am." This immediately cleans up your data and stops the self-bidding.

Then, you apply the same logic down the funnel. Your MoFu campaign, which retargets general website visitors, must exclude your BoFu audience. Why would you show a generic "learn more" ad to someone who has already added a product to their cart? They don't need to learn more; they need a push to complete the purchase. So, your MoFu ad sets must exclude custom audiences like "Added to Cart (14 Days)" and "Initiated Checkout (14 Days)".

And finally, all your campaigns—ToFu, MoFu, and BoFu—must exclude your list of past purchasers. There's little point in paying to acquire a customer you already have (unless you're running a specific campaign for repeat purchases, which should be its own seperate, targeted effort).

This might sound like a lot of work, but you set up these custom audiences once, and then you can reuse them in every new campaign. It takes maybe 20 minutes to build them properly, and it will save you thousands of pounds in wasted ad spend. It turns your account from a chaotic free-for-all into a disciplined, efficient machine. This isn't an advanced trick; it's fundamental. And it's shocking how many businesses, and even some agencies, neglect it. They just keep adding new ad sets and wondering why their performance is so unpredictable and their costs keep creeping up. Now you know why.

This is the main advice I have for you:

I know this is a lot to take in, and it's a very different way of thinking about campaign management than just letting CBO do its thing. The platform makes it seem easy, but to get real, scalable results, you need to impose a logical structure on it. To make it simpler, I've broken down my core recommendations into a clear, actionable table for you to follow.

The bottom line is this: stop thinking of your ad sets as a random collection of audiences and start thinking of them as components in a system designed to move a user from stranger to customer. When you build that system logically, with clean pipes and no leaks between the stages, the problem of rising CPCs due to self-competition simply disappears.

This is precisely the kind of strategic overhaul that can be difficult to execute when you're also trying to run your business. It requires a deep understanding of the auction mechanics and a disciplined approach to account structure. This is where expert help can make a monumental difference. We spend all day, every day, inside ad accounts, diagnosing these exact issues and implementing structures that save our clients money and allow them to scale predictably.

If you'd like to go over your specific account and see how these principles could be applied directly to your campaigns, I'd be happy to offer you a free, no-obligation 20-minute strategy session. We can share screens, look at your setup together, and I can give you some more tailored advice on the spot.

Regards,

Team @ Lukas Holschuh