Solved: Restructure Ad Spend for Better ROAS

Hi there,

Thanks for reaching out and sharing the details of your campaign. It’s a classic situation and one that trips up a lot of people, so don't worry, you're not alone. It's actually a good sign that you're questioning the results instead of just accepting the high purchase volume at face value.

I'm happy to give you some initial thoughts and guidance on this. The short answer is that you're right to be suspicious of that "winning" ad, and you absolutely need to restructure things to give your profitable ads a proper chance to scale. Let's get into how you can fix this.

Let's first address the CBO dilemma...

What you're seeing is Campaign Budget Optimisation (CBO) working exactly as it was designed to, but maybe not how you *want* it to. Meta's algorithm is a machine built to achieve an objective at the lowest possible cost. Your objective is 'Purchases'. So, CBO is laser-focused on one thing: getting you the most purchases for your budget.

The problem is, the algorithm doesn't inherently understand profitability. It just sees that your original ad is very, very good at converting a certain type of user, likely those who are easily persuaded, impulse buyers, or perhaps just cheaper to reach. It's found the "low-hanging fruit." It will pour the majority of the budget there because it’s the path of least resistance to hitting the 'Purchase' objective.

Your other ads, the ones with the much better ROAS, are probably converting a different, more considered type of buyer. These users might be more expensive to reach or take a bit more convincing, so the cost per purchase is higher initially. CBO sees this higher initial cost and thinks, "Why would I spend money there when I can get cheaper purchases over here?". It's a flaw in the logic if your goal is profit, not just volume. You're essentially paying Meta to find you your least profitable customers.

This is a common trap. People see one ad getting all the sales and assume it's the star performer, but they ignore the fact it’s leaking money on every conversion. Your instinct that something is wrong is bang on the money.

That 'winning' ad is a trap... and you need to escape it

Let’s be brutally honest. An ad with a "lousy ROAS" isn't a winning ad. It's a losing ad disguised in vanity metrics. It's like a salesperson who brings in a million quid in revenue but costs the company two million in expenses to do it. They'd be fired, not celebrated. It's the same with your ads.

Forget the number of purchases for a second. Forget the frequency. The only metric that truly matters for an eCommerce business is profitability. Is the ad making you more money than you're spending on it, after accounting for your cost of goods?

Let's visualise this. Imagine Ad A is your current "winner" and Ad B is one of your new, high-ROAS ads that isn't getting much spend.

As you can see, CBO would look at Ad A and see 20 sales versus Ad B's 5 sales and pour all the money into Ad A. But Ad A is barely breaking even (or even losing money, depending on your product margins), while Ad B is a profit-generating machine, even with a tiny budget. Your job is to overrule the algorithm and manually shift the budget to Ad B. You have to kill your darlings, especially when they're costing you money.

I remember one of our eCommerce clients, a women's apparel brand, was in a similar spot. Their account was dominated by one ad that got loads of "Add to Carts" and a handful of cheap sales, but the ROAS was terrible. After we switched it off and isolated their most profitable (but lower volume) ads, we achieved a 691% return for the account. It felt counterintuitive to them at first, but the numbers don't lie.

I'd say you need to force Meta's hand with a new structure

You can't just hope the algorithm figures it out. You need to create an environment where it has no choice but to spend money on your profitable ads. The way to do this is through strategic isolation.

Here’s the exact process I would follow:

1. Turn Off the Losers: Go into your current campaign and turn off the original ad set with that low-ROAS "winning" ad. Don't delete it, just pause it. You're stopping the bleed immediately. Also, pause any other ads or ad sets that aren't delivering a profitable ROAS. Be ruthless.
2. Create a "Winners" Campaign: Duplicate your current CBO campaign. Name the new one something like "[Product] - CBO Winners - [Date]". This keeps things organised.
3. Isolate the Winners: In this new duplicated campaign, go in and delete every single ad set and ad *except* for the ones that have proven to have a good ROAS. You might be left with just one ad set containing 1-3 ads. That's perfectly fine. In fact, it's ideal.
4. Set a Confident Budget: Give this new CBO campaign a healthy budget. At least what you were spending before, if not a bit more. Now, CBO has no other option. It can only spend this budget on your proven, profitable ads.
5. Let It Run: Don't touch it for at least 3-5 days. Let the algorithm stabilise and optimise within this new, constrained environment. It needs time to find the right pockets of the audience for your profitable ads.

This process removes the temptation for the algorithm to go after the cheap, low-quality conversions. You're forcing it to learn how to find more of the high-value customers that your profitable ads appeal to. This is how you scale profitability, not just revenue.

To make it clearer, here’s a simple decision-making flowchart for how you should manage your ads going forward.

You probably should define what "lousy ROAS" actually means

This brings us to a more strategic point. You said the ROAS is "lousy," but what does that mean for your specific business? What is your break-even ROAS? If you don't know this number, you're flying blind. A 2.5x ROAS might be fantastic for a business with 80% margins, but it would be a disaster for one with 30% margins.

To really get a grip on this, you need to understand your Customer Lifetime Value (LTV). How much profit does an average customer generate for you over their entire relationship with your business? Once you know your LTV, you can determine how much you can afford to spend to acquire a customer (your Customer Acquisition Cost, or CAC).

A healthy business model typically aims for an LTV to CAC ratio of at least 3:1. This means for every £1 you spend to acquire a customer, you should get at least £3 back in lifetime profit.

Let's calculate a simplified LTV for your business. You'll need three numbers:

• Average Revenue Per Account (ARPA): What's the average order value? If customers buy multiple times, what's the average they spend per month or year?
• Gross Margin %: After accounting for the cost of the goods you sell, what percentage of the revenue is profit?
• Monthly Churn Rate %: What percentage of customers do you lose each month? (For one-off purchase businesses, this is harder, so you can use 'Repurchase Rate' instead to estimate).

The calculation is: LTV = (ARPA * Gross Margin %) / Monthly Churn Rate

This formula gives you a powerful number. For example, if your LTV is £300, a 3:1 ratio means you can afford to spend up to £100 to acquire that customer. This £100 is your target CAC. Your ROAS target then depends on your Average Order Value (AOV). If your AOV is £50, you need a ROAS of at least 2.0x just to break even on that first purchase (£100 CAC / £50 AOV = 2.0x). To be healthily profitable (3:1), you would aim much higher.

To make this easier, here’s an interactive calculator you can use to get a rough idea of your LTV.

You'll need a better ad testing process

Once you've isolated your winners and they're running smoothly, your job isn't over. Now you need to figure out *why* they're winners and use that information to create your next batch of test ads.

Look at your profitable ads. What's different about them?

• The Angle/Hook: Are they talking about a specific problem? A specific benefit? The low-ROAS ad is probably using a very generic, broad message. The high-ROAS ads are likely speaking to a specific pain point that resonates with a more valuable customer segment.
• The Creative Format: Is it a video? A user-generated style clip? A clean graphic? A carousel? The format itself can filter for different types of buyers.
• The Copy: Read the headline and the primary text. Is it short and punchy or long and detailed? Does it use a question? A bold statement?

Your next step is not to come up with completely random new ideas. It's to iterate on what's already working. Take your best-performing ad and create 3-5 variations of it. Change only one thing at a time: test a new headline, a different first three seconds of the video, or a new image. Launch these new variations in a separate testing campaign (or ad set within your main campaign, with a controlled budget) and run them against your current winner. This disciplined process of iteration is how you find consistent, scalable success instead of just getting lucky once in a while.

This whole proces of optimising campaigns can be time-consuming and requires a fair bit of expertise to get right. It's not just about pushing buttons in Ads Manager; it's about understanding business metrics, customer psychology, and building a rigourous testing methodology.

I've detailed my main recommendations for you below:

Trying to manage this all yourself can be a steep learning curve, with a lot of wasted ad spend along the way. Having an expert eye on your account can help you avoid these common pitfalls, accelerate your learning, and scale your profitable campaigns much faster.

If you'd like to have a more in-depth look at your account and strategy together, we offer a completely free, no-obligation 20-minute consultation. We could walk through your campaigns, identify the biggest opportunities, and give you a clear plan of action. Feel free to book a time that works for you if that sounds helpful.

Either way, I hope this detailed breakdown gives you the clarity and confidence to take control of your campaigns and start scaling profitably.

Hope this helps!

Regards,

Team @ Lukas Holschuh