Solved: Algorithm or Manual Targeting for Facebook Ads?

Hi there,

Thanks for reaching out! It's good you're asking these questions. The dilemma you're facing – whether to trust the algorithm or take the reins yourself – is one of the most common and critical points in scaling an ad account. It's where a lot of people go wrong, either by interfering too much or by trusting blindly.

I'm happy to give you some initial thoughts on this. The short answer is that you're asking a slightly wrong question. It isn't a simple choice between "letting the algorithm run" and "targeting on your own". A properly structured account should be doing both at the same time, systematically. Let me walk you through how I'd approach this to move you from breaking even to being properly profitable.

Your ICP is a Nightmare, Not a Demographic

Before we even get into the ads, let's address a foundational issue. You've identified an age group that converts well. That's a great start, but it's a demographic signal, not a customer profile. Relying on age alone is a trap that leads to generic ads and limited scale. As one of my mentors used to say, your Ideal Customer Profile (ICP) isn't a person; it's a problem state.

Who are these people in that age group? What specific, urgent, expensive problem does your product solve for them? Your $243 conversion didn't happen because of the customer's age; it happened because that person had a problem, and your product was the solution they were looking for at that moment. Your new video ad that "speaks more to this age group" is a good instinct, but it'll be a hundred times more powerful if it speaks directly to their core frustration, their 'nightmare scenario'.

For example, if you're selling productivity software, the nightmare isn't 'being disorganised'. It's the feeling of dread on a Sunday night, knowing you're walking into a chaotic week with missed deadlines and a frustrated boss. Your ads need to hit that nerve. Forget "Save time with our new app." Try something like, "Another Sunday ruined by work anxiety? Stop fighting fires and start preventing them. This is how you get your weekends back."

Once you define the problem, you can find your audience much more effectively than just using age. Where do people with this problem hang out online? What publications do they read? What tools do they already use? This is the work that turns a break-even campaign into a profit machine, long before you even touch the Meta Ads Manager.

I'd say you need a proper campaign structure, not just a better ad set

Right now, you're operating with a single ad set. This is like trying to build a house with only a hammer. It's a tool, but you need a full toolkit and a blueprint. In paid ads, that blueprint is a funnel-based campaign structure. This seperates your audiences based on how familiar they are with your brand and product.

The three core stages are:

• Top of Funnel (ToFu - Prospecting): This is where you find new people who have never heard of you. Your current ad set is a ToFu ad set. This is where you test broad audiences, interest-based targeting, and Lookalike audiences.
• Middle of Funnel (MoFu - Re-engagement): This is for people who have shown some interest but haven't taken a high-intent action. They might have watched one of your videos, visited your website, or engaged with a post. The goal here is to build more trust and move them further down the funnel.
• Bottom of Funnel (BoFu - Retargeting): This is for your hottest audience. People who have visited a product page, added an item to their cart, or started the checkout process but didn't complete it. These ads are direct, often with an incentive, designed to close the deal.

By splitting your campaigns like this, you can tailor your message and budget to each stage. You show different ads to someone who's never heard of you than you do to someone who almost bought from you yesterday. This is how you acheive both scale (with ToFu) and efficiency (with MoFu/BoFu).

You'll need a better prospecting strategy

Let's focus on your ToFu (prospecting) campaign, as that's where your current ad set lives. My advice is to do both. Don't shut down your broad campaign. It’s your exploration tool. The algorithm found a $243 conversion from an unexpected age group; that's gold. It proves the system is capable of finding valuable customers you would have excluded yourself.

Here’s the structure I’d implement immediately:

1. Create one single ToFu Campaign: Name it "CBO - Prospecting - Conversions". Turn on Campaign Budget Optimisation (CBO). This lets Facebook allocate your budget to the best-performing ad set automatically.
2. Keep Ad Set 1 (Your Current Ad Set): Let it continue running with broad targeting in the USA. This is your control and your data-gathering machine.
3. Create Ad Set 2 (Your New Test): This is where you implement your learning. Target the high-performing age group you identified. You can also expand this to the Top 5 English-speaking countries as you planned. Use your new video ad here.
4. Create Ad Set 3 (Future Test): Once you've defined your ICP's 'nightmare', create an ad set based on interests and behaviours that align with that problem. For example, if you sell high-end kitchen knives, you might target people interested in 'Le Cordon Bleu', 'Michelin Guide', and followers of famous chefs, layered with high-income demographics.

Now you have a proper testing framework. You're giving the algorithm freedom in one ad set while applying your human intelligence and learnings in another. CBO will naturally start pushing more budget towards whichever ad set is delivering conversions more efficiently. This structure lets you learn and scale at the same time.

Here's how that might look in a table:

You probably should look beyond Cost Per Conversion

You mentioned you're "breaking even" after being down $1k. This tells me you're focused on the immediate return. While that's understandable, it's a very limiting view. The most successful advertisers I know are obsessed with a different metric: Lifetime Value (LTV). How much is a customer worth to you over their entire relationship with your business, not just the first sale?

Let's do some quick maths. The real question isn't "How low can my Cost Per Conversion go?" but "How high a cost can I afford to acquire a truly great customer?"

LTV is calculated roughly as: (Average Revenue Per Customer * Gross Margin %) / Churn Rate

Once you know your LTV, you know what you can afford to spend on customer acquisition cost (CAC). A healthy ratio is often cited as 3:1 (LTV:CAC). So if your LTV is £900, you can afford to spend up to £300 to acquire a customer and still have a very profitable business model. Suddenly, paying £50 or even £100 for a conversion doesn't seem so scary, does it? It looks like an investment.

This is the mindset that unlocks aggressive, intelligent scaling. It frees you from the tyranny of trying to get the cheapest possible leads and allows you to focus on acquiring the *best quality* customers, who will spend more over time.

Use the calculator below to get a rough idea of your own LTV. Play with the numbers and see how small changes in churn or average revenue can massively impact how much you can afford to spend on ads.

We'll need to look at what 'winning' actually means

You’ve been running ads for a week. In the world of Meta ads, that's barely a moment. The algorithm is still in its "learning phase", trying to understand who your ideal customer is. Making drastic changes now, like killing a campaign, is premature. My rule of thumb is to let an ad set spend at least 2-3x your target CPA before making a judgement call. If your target CPA is $50, an ad set needs to spend $100-$150 without a conversion before I'll even consider turning it off.

Your job is to feed the machine with high-quality inputs (good creative, clear offers) and well-structured tests, then have the patience to let it work. This means letting ad sets run long enough to gather statistically significant data.

Below is a conceptual illustration of what you might see after running the structured test for a couple of weeks. You might find that your new age-targeted ad set has a lower Cost Per Acquisition (CPA), but the broad ad set, while more expensive per conversion, is the only one that delivers those high-value customers like your $243 sale. This is data you can act on. Maybe you scale the age-targeted ad set for volume, while keeping the broad one running to hunt for whales.

Ultimately, a successful paid advertising strategy isn't about finding one magic bullet ad set. It's about building a robust system that allows for continuous testing, learning, and optimisation across the entire customer journey. I've detailed my main recommendations for you below:

This approach might seem more complex, but it's how you build a scalable and resilient advertising engine. It protects you from platform changes and gives you clear levers to pull to grow your bussiness. It takes time and expertise to manage this kind of structure effectively, which is why many businesses choose to work with a specialist.

We help businesses implement these kinds of advanced strategies every day. If you'd like to have a more in-depth chat and have us review your ad account in real-time, we offer a completely free, no-obligation 20-minute strategy session. We could walk through your exact setup and provide some more specific, actionable advice.

Hope this helps!

Regards,

Team @ Lukas Holschuh