Solved: Best Way to Test Facebook Ads Without Wasting Money

Hi there,

Thanks for reaching out. Your questions about testing ads highlight a frustration I see a lot. People get told a dozen different 'right ways' to do it, and most of them end up just burning cash while Meta's algorithm has a party on their budget.

I'm happy to give you some initial thoughts. The truth is, you're asking good tactical questions, but you might be focusing on the wrong part of the problem. Your testing method isn't the real issue. We need to look a bit deeper at the strategy behind the ads first, and then build a testing framework on top of that that's based on maths, not myths. Let's get into it.

The Real Reason Your Ads Are Burning Money (And It's Not CBO)

You're stuck in what I call the 'testing trap'. You're tweaking the knobs on the machine – CBO, ABO, budget allocation, number of creatives – hoping one combination will magically unlock a flood of sales. But the machine can only work with the fuel you give it. If the fuel is bad, it doesn't matter how expertly you tune the engine; you're going nowhere.

The fuel, in this case, is your understanding of your customer and your offer. Before you spend another pound on testing which shade of blue a button should be, you need to answer one question with brutal honesty: What is my customer's nightmare?

Forget the demographic profiles. "Women aged 25-40 who like yoga" tells you nothing. That describes millions of people with wildly different motivations. Your Ideal Customer Profile (ICP) isn't a person; it's a problem state. A deep, urgent, and expensive problem that keeps them awake at night.

• -> For a company selling project management software, the nightmare isn't 'needing to organise tasks'. It's the CTO being terrified of a critical product launch failing because communication is a mess and deadlines are being missed, putting his reputation on the line.
• -> For an e-commerce brand selling premium skincare, the nightmare isn't 'dry skin'. It's the feeling of losing confidence and feeling invisible as you get older, watching competitors who seem to have it all figured out.
• -> For a service provider selling financial advice, the nightmare isn't 'wanting to save for retirement'. It's the crippling anxiety of looking at your bank account and realising you're one unexpected bill away from disaster, with no clear path forward.

Your ads, your creatives, your entire message must speak directly to that nightmare. When you do that, your ads stop being an interruption and start being a solution. Your creative testing becomes about finding the most potent way to articulate that solution, not just throwing random images at the wall to see what sticks. Most businesses fail at this stage, which is why their ad accounts are a graveyard of failed tests.

The whole process should look less like a random experiment and more like a logical progression. You start with the core problem, and everything else flows from there.

You probably should stop looking at the wrong metrics...

Let's talk about your metrics. You mentioned a CPC of $8+ and a CTR under 1.5%. You've been trained to think these are 'bad' numbers. I'm here to tell you that, in isolation, they mean absolutely nothing.

A high CPC or a low CTR are diagnostic metrics. They can sometimes hint at a problem (e.g., your creative isn't grabbing attention, or your audience isn't right), but they are not decision-making metrics. You could have a $20 CPC, but if every click results in a £5,000 sale, you'd be a fool to turn that ad off. Conversely, you could have a £0.10 CPC, but if 10,000 clicks result in zero sales, you've just found a very efficient way to go broke.

The only metric that truly matters for making decisions is your Cost Per Acquisition (CPA), and more importantly, how that CPA stacks up against your Customer Lifetime Value (LTV).

The real question isn't "How low can my CPA go?" but "How high a CPA can I afford to acquire a great customer?" Once you know that number, you are freed from the tyranny of vanity metrics. You can advertise with confidence, knowing exactly when a campaign is a profitable investment and when it's a money pit.

So, how do you figure this out? We need to do some simple maths. It's the most important calculation you will ever do for your business.

1. Average Revenue Per Account (ARPA): How much revenue does a typical customer bring in per month (or year, just be consistent)?
2. Gross Margin %: What's your profit margin on that revenue after accounting for cost of goods sold?
3. Monthly Churn Rate %: What percentage of customers do you lose each month?

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

Let's say you run a subscription box service. ARPA is £50/month, your gross margin is 70%, and you lose 5% of your customers each month.
LTV = (£50 * 0.70) / 0.05 = £35 / 0.05 = £700.

Each customer is worth £700 in gross margin to your business. A healthy LTV to Customer Acquisition Cost (CAC) ratio is typically 3:1. This means you can afford to spend up to £700 / 3 = ~£233 to acquire a single customer and still have a very healthy, scalable business. Your Target CPA is £233.

Now you have your weapon. Now you know the real benchmark. An ad isn't 'failing' because its CPC is high; it's failing if it spends more than £233 without getting you a new subscriber. This is the maths that unlocks intelligent growth.

To make this easier, I've built a calculator for you. Play around with the sliders to see how small changes in your business model can dramatically affect how much you can afford to spend on ads.

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

Okay, now that we've established the strategic foundation – know your customer's nightmare and know your numbers – we can finally build a testing framework that actually works. This is where we answer your specific questions about CBO vs. ABO.

The problem you described with CBO is classic. You put 3-4 creatives in an ad set, and Meta's algorithm almost instantly picks one and starves the others of budget. It does this because its only goal is to get you the cheapest results based on very early, often misleading, data. It's not trying to find your *best* creative long-term; it's trying to find the *easiest* win right now. This makes CBO a terrible tool for fair, controlled creative testing.

Here’s the simple, correct way to structure things:

Phase 1: Testing -> Ad Set Budget Optimization (ABO)

When you want to test creatives (or audiences) against each other fairly, you need to control the spend. ABO is the tool for this. You set the budget at the ad set level, forcing Meta to spend that amount on that specific ad set.

• -> Campaign Type: ABO (set budget at ad set level).
• -> Structure: Create one ad set for each creative you want to test. That's right, 1 creative per ad set. If you have 5 creatives to test, you create 5 identical ad sets, with the only difference being the single creative inside each one.
• -> Budget: Give each ad set the same daily budget (we'll discuss how much in the next section).

This setup forces a level playing field. Every creative gets an equal amount of budget to prove itself. There's no algorithm favouritism. The winner is decided by pure performance based on your key metric (CPA), not by which ad got lucky in the first few hours. This is the strategy we use to find winning creatives for our clients, like in one e-commerce campaign where we achieved a 691% return for a women's apparel brand by systematically testing creative concepts this way.

Phase 2: Scaling -> Campaign Budget Optimization (CBO)

Once your ABO testing campaign has identified a clear winner (or two), that's when you bring in CBO. CBO is a scaling tool, not a testing tool.

• -> Campaign Type: CBO (set budget at the campaign level).
• -> Structure: Create a new campaign. Inside it, create ad sets for your best-performing *audiences*. Then, put your 1-3 proven *winning creatives* from the testing phase into each of these ad sets.
• -> Budget: Set one single, larger budget at the campaign level.

Now you let Meta's algorithm do what it's good at. It will automatically shift the budget between your proven audiences to find the cheapest conversions, all while using your best-performing ads. You've done the hard work of finding the winners in a controlled environment; now you're just letting the algorithm efficiently scale them.

I'd say you need a kill switch based on maths, not feelings...

Now for the million-dollar question: "How soon should I kill a creative?" You've heard you need to give it a full 3 days. That is one of the most persistent and destructive myths in paid advertising. It's completely arbitrary.

It is not about time; it is about spend.

An ad doesn't need "time" to learn. It needs data. Data comes from impressions, clicks, and conversions, which are all a function of how much you spend. Waiting three days on a £5/day budget is pointless. You'll have almost no data. Conversely, on a £500/day budget, you might have all the data you need to make a decision in a few hours.

Here is your new rule. It's simple, data-driven, and it will save you a fortune. You learned how to calculate your Target CPA in the previous section. Let's say it's £50.

The Kill Switch Rule: Turn off any ad or ad set once it has spent 1.5x your Target CPA without a single conversion.

Some people are more aggressive and use 1x, some are more conservative and use 2x. I find 1.5x is a good middle ground. So in our example, once an ad set has spent £75 (1.5 * £50) and has not generated a purchase, you kill it. Without emotion. It has had its chance to perform and failed. The maths have spoken.

This answers your other questions too:

What's the best budget per creative?
It depends entirely on your Target CPA. A good rule of thumb is to set your daily budget per ad set (which contains one creative in our testing setup) to be high enough to hit your 1.5x CPA kill switch within 3-4 days. This gives the ad a reasonable amount of time to gather data without dragging the test on for ever.
So, if your Target CPA is £50, your "kill spend" is £75. A daily budget of £20-£25 per ad set would be perfect. It'll hit the kill switch in 3-4 days if it's a dud.

Is 5 ad sets at $15/day spreading the budget too thin?
Again, it depends on your Target CPA. If your Target CPA is, say, $30, then your kill spend is $45. A $15/day budget is fine; you'll make a decision in 3 days. But if your Target CPA is $150, your kill spend is $225. At $15/day, it would take 15 days to reach that spend! Yes, in that case, you are definitly spreading the budget too thin, and the test will take forever. You'd be better off testing just 2-3 creatives at a higher daily budget to get a faster result.

Here's a table to make it clearer:

I've detailed my main recommendations for you below:

This has been a lot of information, I know. It's a fundamental shift from tweaking tactics to building a proper marketing strategy. To summarise, here is the exact plan I would put in place if we were working together. This is the main advice I have for you:

Implementing this framework requires discipline. You have to trust the maths and be ruthless about cutting what isn't working. But this is how you move from gambling on ads to building a predictable, scalable customer acquisition engine. It’s how we've been able to scale accounts, like one software client for whom we generated over 5,000 trial signups at a consistent $7 cost per trial.

This might seem like a lot to take in and implement all at once, and honestly, it is. Getting this right involves not just setting up the campaigns correctly, but also deeply understanding your market, your numbers, and how to interpret the data as it comes in. It's a continuous process of analysis and optimisation.

If you'd like to walk through this on your own ad account and build a specific testing plan for your business, we offer a completely free, no-obligation 20-minute strategy session. We can look at your current setup, help you calculate your core numbers, and lay out the first steps of a robust testing framework. It’s often the fastest way to get clarity and stop burning money.

Regards,

Team @ Lukas Holschuh