Solved: Facebook Ads Spend Concentrated on Few Ads

Hi there,

Thanks for reaching out! Happy to give you some of my initial thoughts and a bit of guidance based on what you've described. It's a really common situation to find yourself in with Meta ads, so don't worry, you're not alone in seeing this.

Let's get right into it.

We'll need to look at your current ad setup...

First off, your direct question about the ads that are only spending pennies. The short and brutally honest answer is yes, you should probably kill them. When Facebook's (or Meta's) algorithm decides to put almost all the daily budget behind just one or two ads in a campaign, it's sending you a very clear signal. It has already determined, based on its vast data and initial user reactions, that those specific ads are the most likely to get you the results you're optimising for. The other ads, for whatever reason, are seen as weaker performers.

Letting them run "until they reach a certain amount of impressions" is often a waste of time and can actually drag down your overall campaign performance. If the algorithm does eventually decide to give them a bit more spend and they get impressions but no clicks or conversions, it just muddies your data and lowers your campaign's quality score. It's better to be decisive. The algorithm has essentially done the initial split test for you and picked a winner. Trust that signal.

It's perfectly normal for this to happen. I'd say in 9 out of 10 campaigns we run, the budget naturally gravitates towards a small subset of the ads. What you can do, and what we often do for clients, is take your winning ad (the one getting all the spend and results) and duplicate it into a new, fresh campaign. Sometimes you can even duplicate the whole ad set. This isolates the top performer and gives the algorithm a clean slate to work with, focused entirely on what you know already works. Then you can confidently turn off the original campaign with all the underperforming assets.

However, this is really just treating a symptom. The bigger question isn't just which ad is working, but why it's working and how we can find more audiences and strategies that work just as well, if not better. This leads us onto the more fundamental part of your advertising strategy: your campaign structure and targeting. This is where most people go wrong and where the biggest gains are usually found. I remember one client, a medical job matching SaaS, that was struggling. We reduced their Cost Per User Acquisition from £100 to £7 by restructuring their account and refining their targeting. The ads themselves didn't change as much as the strategy behind who was seeing them.

I'd say you need a proper funnel structure...

When I audit Meta accounts, probably the single biggest issue I see is a lack of clear structure. People often throw a bunch of different audiences into one campaign, or test audiences that don't really align with their overall goals. To get consistent results and be able to scale effectively, you need to think in terms of a funnel: Top of Funnel (ToFu), Middle of Funnel (MoFu), and Bottom of Funnel (BoFu).

This means structuring your campaigns around the customer's journey, not just lumping everything together. Here's how I'd typically prioritise the audiences you should be testing, moving from the coldest audiences to the warmest, most valuable ones.

META (Facebook/Instagram) ADS AUDIENCE PRIORITISATION

ToFu (Top of Funnel - Reaching New People)

These are your cold audiences. People who have never heard of you before. The goal here is to introduce your brand and get them to your website or landing page.

• 1. Detailed targeting (interests, behaviours, demographics): This should be your starting point for any new account. It's how you gather the initial data you need for everything else to work.
• 2. Broad targeting: This is where you target a country and maybe an age/gender, but leave interests blank. It sounds crazy, but once your account has enough conversion data (your "pixel is seasoned"), Meta's algorithm can be incredibly effective at finding customers for you. You only do this once you have hundreds, ideally thousands, of conversions tracked.
• 3. Lookalike audiences: These are audiences Meta builds that "look like" a source audience you provide. This is powerful, but the quality of your lookalike depends entirely on the quality of your source. I'd prioritise them in this order:
  • -> Lookalike of highest value previous customers
  • -> Lookalike of all previous customers
  • -> Lookalike of people who have purchased
  • -> Lookalike of people who added payment info
  • -> Lookalike of people who initiated checkout
  • -> Lookalike of people who added to cart
  • -> Lookalike of all website visitors
  • -> Lookalike of people who watched 50% of your video ads

MoFu (Middle of Funnel - Re-engaging Interested People)

These are people who have shown some interest but haven't taken that final step. You're retargeting them to bring them back.

• -> People who visited your landing or product page (but exclude anyone who purchased or reached the checkout pages)
• -> People who viewed your product catalogue (again, excluding recent buyers)
• -> All website visitors (a broader group, but still valuable)
• -> People who watched a significant portion (e.g., 10 seconds or 50%) of your video ads

BoFu (Bottom of Funnel - Closing the Deal)

These are your hottest prospects. They were on the verge of converting. The message here should be about overcoming final objections, perhaps with an offer or a reminder.

• -> People who added payment info
• -> People who initiated checkout
• -> People who visited the checkout page
• -> People who viewed their cart
• -> People who added items to their cart

And finally, you have your existing customers, who you can target for repeat purchases or to upsell other products.

This might look complicated, but the logic is simple: the further down the funnel someone is, the more likely they are to convert. So you prioritise your budget and testing accordingly. If you have a small budget, you can combine MoFu and BoFu audiences into a single "retargeting" campaign, but you should always keep them separate from your cold ToFu campaigns. This structure gives you clarity on what's working at each stage of the journey.

You probably should be more specific with your targeting...

Having the right structure is one thing, but filling it with the right audiences is another. This is where that expertise and experience really comes into play. It's not just about picking interests that sound right; it's about understanding the nuances of the platform.

For new accounts, you have to start with detailed targeting to get that initial data. You need at least 100 people in a custom audience (like 'website visitors') before you can even use it for retargeting, and honestly, you need way more than that for it to be effective. So detailed targeting is your first step.

Now, the biggest mistake I see people make here is choosing interests that are far too broad. I remember a client who was trying to sell a service to eCommerce store owners. They were targeting the interest "Amazon". On the surface, it might seem logical. But think about it. Who is in that audience? Millions of people who shop on Amazon, people who work for Amazon, people who follow Amazon news... the vast majority are not your target customer. It's a hugely inefficient way to spend your money.

The key is to find interests that your target audience is much more likely to have than the general population. So for that eCommerce client, we switched the targeting to things like:

• -> People who are admins of a Facebook 'Retail Page'.
• -> Interests in eCommerce software like 'Shopify', 'WooCommerce', or 'BigCommerce'.
• -> Interests in other tools eCommerce owners use, like 'Klaviyo' or 'Mailchimp'.
• -> Interests in eCommerce-focused publications or influencers.

See the difference? You're honing in on signals that are much more specific to your ideal customer. Someone interested in 'Shopify' is far more likely to be an eCommerce store owner than someone just interested in 'Amazon'. This is the sort of strategic thinking you need to apply to your own niche. Think about the specific tools, publications, events, or brands that only your ideal customer would be interested in.

Once you have enough data from this kind of specific interest targeting (we're talking at least 100 conversions, but ideally 500-1000), you can then move on to building those powerful lookalike audiences. And again, be specific. A lookalike of 'all website visitors' is okay, but a lookalike of 'purchasers' is much, much better. A lookalike of your 'highest value customers' is better still. Always start with the lookalike audience that is based on the most valuable action a user can take. This is how you find more people like your best customers, which is the whole point of scaling.

Finally, you have to test. And test properly. With a ToFu/MoFu/BoFu structure, you have separate campaigns for each stage. Within your ToFu campaign, you can have multiple ad sets, each testing a different audience (e.g., one for Shopify interests, one for a lookalike of purchasers, etc.). Let them run for a few days and see which one performs best. A good rule of thumb is to turn off any ad set that has spent about 2-3 times your target cost-per-acquisition (CPA) without getting a single conversion. If your target CPA is £20, and an ad set has spent £60 without a result, it's very unlikely to become a winner. Be ruthless, kill the losers, and reallocate that budget to the winners or to testing new audiences. I remember working with a client that sold cleaning products where we increased their revenue by 190% and acheived a 633% return by constantly testing new interest groups and lookalikes within a clear funnel structure.

You'll need realistic expectations for your costs...

A question I get asked all the time is "What should my cost per conversion be?". The honest answer is: it depends. It's affected by your industry, your targeting, your offer, the quality of your ads, and a dozen other things. However, based on our experience running hundreds of campaigns, we can give you some rough ballpark figures.

Let's break it down. Your Cost Per Acquisition (CPA) is basically your Cost Per Click (CPC) divided by your Landing Page Conversion Rate (CVR). So, to estimate your CPA, we need to estimate those two numbers first.

In developed countries (UK, US, Canada, Australia, Western Europe etc.), a typical CPC on Meta ads might be in the range of £0.50 to £1.50. In developing countries, it can be much lower, maybe £0.10 to £0.50, but the quality of the traffic is often lower too. For a simple conversion like a lead form fill, an email signup, or a free trial, a decent landing page might convert between 10% and 30% of the traffic it gets.

Let's do the maths for a lead/signup objective in a developed country:

• Best-case scenario: £0.50 CPC / 30% CVR = £1.67 per signup
• Worst-case scenario: £1.50 CPC / 10% CVR = £15.00 per signup

So, a realistic range is somewhere between £1.60 and £15 per lead. If you're getting leads for £4, for example, that's actually pretty good and falls squarely within that normal range.

Now, if your goal is an actual sale (eCommerce), your conversion rates will be much lower. A typical eCommerce store converts between 2% and 5% of visitors. Let's look at the maths for that:

• Best-case scenario: £0.50 CPC / 5% CVR = £10.00 per purchase
• Worst-case scenario: £1.50 CPC / 2% CVR = £75.00 per purchase

Here's a table to make that clearer:

The goal of all the optimisation work—the testing, the targeting, the funnel structure—is to push your numbers towards that best-case scenario. Better targeting and better ads lower your CPC. A better landing page and offer increases your CVR. It's a constant process of improvement. I recall one campaign we ran for a subscription box client where we got to a 1000% Return on Ad Spend (ROAS) precisely by optimising every single one of these levers over time.

This is the main advice I have for you:

To pull this all together, running sucessful campaigns isn't about finding one magic ad. It's about building a systematic, repeatable process. It's about structure, testing, and continuous optimisation. I've detailed my main recommendations for you below in a more scannable format.

As you can see, it's a bit more involved than just turning ads on and off. It's about building a robust system. It's this system that allows agencies and experts to get repeatable results for clients, whether it's generating 5,082 software trials at $7 each or generating $115k revenue in 1.5 Months for course sales. The principles are the same.

Getting this right takes time, experience, and a deep understanding of how the platforms work. It's not just about setting up the ads; it's about the ongoing analysis, strategic adjustments, creative testing, and landing page optimisation that make the real difference between a campaign that breaks even and one that is wildly profitable.

If you'd like to go over your specific situation in more detail, we offer a free, no-obligation initial consultation where we can review your account and strategy together. It can be incredibly helpful to get a second pair of expert eyes on things and develop a clear path forward.

Hope this helps give you a much clearer picture of how to approach things!

Regards,

Team @ Lukas Holschuh