Solved: Winning Facebook Ad Creative Dominating Budget

Hi Matt,

Thanks for reaching out. It's good to hear you're seeing some solid performance, finding a winning creative that keeps improving is a great place to be in. A lot of people struggle to even get that far.

You’ve asked about testing new creatives and how to do it without your current winner soaking up all the budget. It's a really common question, and your approach of setting up a new adset is what a lot of people do. But to be honest, it points to a bigger, more fundamental issue with how most people think about running ads on Meta. I’m happy to give you a few of my thoughts on it, which might be a bit different from what you've heard before.

The short answer is you're not really 'missing' something simple, but you might be focusing on the wrong part of the problem. You're trying to solve a creative testing issue, but the real path to scaling isn't just about finding the 'next' winning video. It's about building a machine that finds winning audiences and lets you systematicaly test your messaging to them, and that's a whole different ball game.

The problem with chasing the next 'winning creative'...

First off, let’s talk about why your winning creative is taking all the budget. That’s not a bug, it’s a feature. When you put multiple ads in one adset, Facebook’s algorithm is designed to do exactly that: find the ad it predicts will get you the best results (in your case, purchases) for the lowest cost, and then it pushes the majority of the budget to that ad. It’s doing its job. It's found a winner and its exploiting it for all its worth.

So when you add new creatives to that same adset, they are almost always going to struggle to get any spend. They are unproven. The algorithm looks at your established winner with all its performance history and says, "Why would I risk the budget on this new thing when I know this old one works?". This is why your test in a separate adset got some results - you forced Facebook to spend the money on it. It had no other choice.

But here’s the contrarian bit. Constantly hunting for the next winning creative is a bit of a fool's errand. It turns advertising into a lottery where you're just hoping the next video you make is a hit. A truly scalable ad account isn't built on one-off creative wins; it's built on a deep understanding of your customer, rigorous audience testing, and a solid grasp of your business's numbers. The creative is just the vehicle for a message that should already be proven to resonate with a specific audience.

I remember one eCom client who came to us with the exact same issue. They had a "hero" video ad that did really well, and they spent months trying to replicate it, making dozens of variations. Their results were all over the place. We stopped the endless creative churn and instead focused on restructuring their account around audiences. The result was a 1000% return on ad spend for their subscription box. The creative was important, sure, but it was the targeting that unlocked the real growth.

We'll need to look at a proper testing framework...

Okay, so if just chucking new ads into an adset isn't the way, what is? You need a structure. Your current method of a new adset at £15/day is a start, but it's not statistically sound. You got a £2.20 purchase straight away, which is great, but it could easily have been luck. If you'd run it for another day and got no sales, your CPA would have shot up. You can't make reliable decisions based on such small amounts of data. It’s like flipping a coin once, getting heads, and declaring it a double-headed coin.

Here’s a more robust way to think about it.

Dynamic Creative Testing (DCT)

Instead of creating a new adset for each new video, you should be using Facebook's Dynamic Creative feature. This is what it's built for. You give it a bunch of separate components:

• -> Multiple Videos/Images
• -> Multiple Headlines
• -> Multiple Primary Texts (the copy above the ad)
• -> Multiple Descriptions

Facebook then mixes and matches all these components to find the best-performing combinations. It does the testing for you, far more efficiently than you can do it manually. This tells you not just which video works, but which message (headline/copy) works with which video. It gives you much richer data.

Campaign Budget Optimisation (CBO) vs. Adset Budget Optimisation (ABO)

You mentioned you're using an adset budget (£50/day on your main one). This is fine, but for testing, CBO can often be better. With CBO, you set the budget at the campaign level, and Facebook distributes it across your adsets to the ones it thinks will perform best. This is really powerful for testing audiences, which we'll get to next.

A simple testing structure could look like this:

Campaign 1: CBO Testing Campaign - £50/day

• -> Adset 1: Audience A (e.g., Broad Interests)
  • --> Dynamic Creative Ad (with your 3-4 new videos + headlines/copy)
• -> Adset 2: Audience B (e.g., Lookalike of Purchasers)
  • --> Dynamic Creative Ad (same one as above)
• -> Adset 3: Audience C (e.g., Different Interests)
  • --> Dynamic Creative Ad (same one as above)

You let this run until each adset has spent enough for you to make a call (a common rule of thumb is at least 1-2x your target CPA). Then you can see which audience performs best, and within that adset, Facebook's breakdown feature will tell you which creative components are working. You kill the losing adsets and scale the winning one, or move the winning audience into its own "proven" campaign.

I'd say you need to think beyond the creative and focus on your audience...

This is probably the most important shift in thinking you can make. You’re asking "How do I test this creative?". The better question is "Who is the right audience for my product, and what creative will work for them?".

The success of your campaigns will be determined far more by your targeting than by your creative. A mediocre ad shown to a perfect audience will always, always beat a brilliant ad shown to the wrong people. You could have the most amazing, cinematic video in the world, but if you show it to people who have zero interest in your product, you'll get zero sales.

When we take on a new account, the very first thing we do is map out the audience structure. Here’s how we prioritise audiences, from coldest to warmest. The goal is to test these systematicaly.

Your current winning adset is likely a ToFu audience. The real, sustainable profit often comes from nailing your MoFu and BoFu retargeting. These are people who are already warmed up. Hitting them with a specific offer or reminder can be incredibly effective and cheap. You should have separate, always-on campaigns for these audiences.

For one of our B2B SaaS clients, we were able to get 4,622 registrations at just $2.38 each. This wasn't because of one magic video. It was because we relentlessly tested different Lookalike audiences combined with retargeting layers until we found the pockets of users who were most likely to convert.

You probably should define your ideal customer by their problem, not their clicks...

Drilling down even further than audiences, you need to get inside the head of your customer. Who are they, really? What keeps them up at night? Why do they need what you sell? Forget demographics for a second. "Women aged 25-40 who like online shopping" is useless.

You need to define your customer by their pain. Their urgent, expensive, frustrating nightmare. Your product is the solution to that nightmare. Once you understand this, your ad creative and copy will write itself.

Let's imagine you sell high-quality, comfortable work-from-home clothing.

The demographic is "people working from home". Useless.

The nightmare is "I have a surprise Zoom call with my boss in 5 minutes and I'm wearing a stained t-shirt from 3 days ago. I feel unprofessional and panicked."

Now, think about the ad you could make for that. It’s not just showing a nice jumper. It’s showing someone in that exact panic scenario, then quickly throwing on your jumper and instantly looking professional and feeling confident. You're not selling clothes; you're selling relief and professionalism. That's a message that cuts through the noise.

This is the work you need to do before you even press record on a new video. Who are you talking to? What is their specific problem? How does your product solve it in a way that makes them feel something (relief, confidence, excitement)?

When you have this, you can target better too. Where do these people hang out online? What brands do they follow? What publications do they read? Those become your detailed targeting interests. This intelligence is the foundation of your entire strategy. Without it, you’re just guessing.

You'll need to understand your numbers properly...

This is the part that separates the amateurs from the pros. You’re happy with a £2.20 cost per purchase. And you should be, that sounds fantastic on the surface! But the real question is: what is a customer actually worth to you?

If you don't know this, you're flying blind. You don't know how much you can actually afford to spend to acquire a customer, which means you don't know when to scale a campaign or when to kill it. This is where calculating your Customer Lifetime Value (LTV) comes in.

Let’s run a hypothetical example. Let's say you sell products that people might buy more than once.

1. Average Order Value (AOV): How much does a customer spend in a typical purchase? Let's say it's £40.
2. Gross Margin %: After the cost of the goods, what's your profit margin? Let's say it's 60%.
3. Purchase Frequency: How many times does the average customer buy from you in a year? Let's say it's 3 times.
4. Customer Lifetime (Years): How long does a customer typically stick with you? Let's say it's 2 years.

Now, we can do some simple maths:

LTV = (AOV * Purchase Frequency * Customer Lifetime) * Gross Margin %

LTV = (£40 * 3 * 2) * 0.60

LTV = £240 * 0.60 = £144

In this example, each new customer you acquire is worth £144 in gross profit to your business over their lifetime. This completely changes the game. A common benchmark is to aim for a 3:1 LTV to Customer Acquisition Cost (CAC) ratio. This means you can afford to spend up to £48 (£144 / 3) to acquire a single new customer and still have a very healthy, profitable business.

Suddenly your £2.20 CAC doesn't just look "good", it looks like an absolute goldmine that you should be pouring money into as fast as you can! It also tells you that even if your CAC on a new test adset creeps up to £10, £20, or even £30, it could still be wildly profitable for you. Knowing your LTV frees you from the tyranny of only chasing ultra-low CPAs and gives you the confidence to scale aggressively.

So, what should you do tomorrow?

That was a lot of information, I know. It can feel a bit overwhelming. The point isn't to change everything overnight. The point is to start thinking like a professional advertiser, not just a guy making ads. It's about building a system, not just finding a winning creative.

I've detailed my main recommendations for you below to give you a clear path forward.

I realise this is a big shift from where you are now. Moving from fiddling with adsets to building out a full-funnel strategy with proper testing protocols and a deep understanding of your unit economics is a significant step up. It's the difference between having a successful ad and having a successful advertising system that can grow your business predictably for years to come.

This is, frankly, where having an expert eye can make a huge difference. We do this stuff day in, day out. We've run these kinds of restructures for dozens of clients, from eCom stores to B2B software companies, and the principles are always the same. For one women's apparel brand, implementing a proper audience and retargeting structure like this led to a 691% return on their ad spend. It's about having a proven process.

If you'd like to have a chat and go through your account together, we offer a free, no-obligation initial consultation. We could pull up your Ads Manager on a screen share and I can show you exactly how we'd apply these concepts to your specific situation. It might help make all of this theory much more concrete.

Either way, you're doing well to be asking these questions. Keep that curiosity up.

Hope that helps!

Regards,

Team @ Lukas Holschuh