Solved: Facebook Ad ROAS Low Despite High Conversions

Hi there,

Thanks for reaching out! Happy to give you some initial thoughts on the situation with your Meta ads. It's a classic, frustrating problem that I see all the time, so you're definately not alone. You've got one ad that the algorithm loves, but its performance is just... meh. Meanwhile, your potentially brilliant ads are getting ignored.

The good news is that this is fixable. It's not about finding a magic button, but about understanding why the algorithm is behaving this way and then taking back control. Let's get into it.

We'll need to look at why Meta's algorithm is a greedy idiot...

Right, first things first. You need to stop thinking of the Meta algorithm as this all-knowing genius. It's not. It's a machine programmed to do one thing: get you the result you asked for (in this case, purchases) as cheaply and quickly as it can. It doesn't care about your profit margins, your ROAS, or the quality of the customer. It's a short-term thinker, and that's the core of your problem.

You said your one winning ad has a "lousy ROAS" but is getting all the spend. This is completely normal behaviour for Campaign Budget Optimization (CBO). Here's what's likely happening under the bonnet:

1. Early Success Bias: That ad probably got a few cheap, easy purchases right at the start. Maybe it hit a pocket of impulse buyers or people who were already close to converting. The algorithm saw this and thought, "Brilliant! This is the one!" and immediately started pumping the majority of your budget into it.

2. The Path of Least Resistance: The algorithm will always, always follow the easiest path to a conversion. Your high-ROAS ads might appeal to a more discerning, higher-value customer. This customer might take a bit longer to decide, or they might be more expensive to reach. The algorithm sees this as 'hard work' and ignores it in favour of the ad that's getting the quick, cheap wins, even if those wins are less profitable for you in the long run.

3. It's Optimising for CPA, Not ROAS: Even though you're looking at ROAS, the CBO's primary driver when you select 'Purchases' is Cost Per Acquisition (CPA). It's trying to get you the lowest CPA possible. Your "lousy ROAS" ad probably has a slightly lower CPA than your other, better ads. Even a difference of a few quid can be enough for the algorithm to decide where to spend 90% of your money. It doesn't have the context to understand that an ad with a £20 CPA that brings in a £100 sale (5x ROAS) is infinitely better than an ad with a £18 CPA that brings in a £30 sale (1.6x ROAS).

You are basically paying Facebook to find you your least profitable customers because they are the easiest to convert. It's a trap many advertisers fall into. They see the purchase number go up and think it's working, without realising their profit is being eroded. The fact that you've spotted the ROAS issue puts you ahead of most.

I'd say you need to take back control of your testing...

So, what's the fix? You can't just shuffle ads around in your existing CBO campaign and hope for the best. That's like trying to reason with a vending machine. You need to change the game entirely. The solution is to separate your testing from your scaling.

CBO is a scaling tool, not a testing tool. You use it *after* you know what works. Right now, you're still figuring that out. So, you need to switch to Ad Set Budget Optimization (ABO) for your testing phase.

Here’s the structure I would recomend:

Campaign 1: The Testing Ground (ABO)

• Objective: Conversions (Purchases)
• Budget Setting: Ad Set Budget Optimization (ABO)
• Structure: Create multiple ad sets. Each ad set should target a different audience you want to test (e.g., one for broad, one for a specific interest stack, one for a lookalike audience).
• Budget Allocation: Give each ad set an equal, small budget. Let's say £10-£20 per day each. This is crucial. It forces Meta to give each of your audiences and ads a fair shot.
• Ads: Inside each ad set, place 3-4 ads. These should be your existing ads, including the high-ROAS ones and some brand new ideas.

You run this ABO campaign for 3-5 days. Don't touch it. Let the data come in. After a few days, you'll have a much clearer picture. You'll be able to see which *combination* of audience and creative is actually profitable. You're not looking for the ad with the most purchases; you're looking for the ad with the best ROAS that has proven it can get consistent sales.

Campaign 2: The Scaling Machine (CBO)

• Objective: Conversions (Purchases)
• Budget Setting: Campaign Budget Optimization (CBO)
• Structure: Once you've identified a winning ad set and ad from your ABO campaign, you duplicate that ad set into this new CBO campaign. Only the proven winners get to play here.
• Budget Allocation: You give this CBO campaign the majority of your budget. Because it only contains proven performers, you can trust the algorithm to scale it more effectively. It might still favour one ad over another, but since they are all profitable, it matters less.

This two-campaign structure is fundamental. You have an always-on 'testing ground' where you're constantly trying new audiences and creatives on a small, controlled budget. And you have a 'scaling machine' where you send your champions to bring home the profits. When a winner in the CBO campaign starts to fatigue (performance drops), you pause it and promote the next winner from your testing campaign. It's a continuous cycle of optimisation that puts you in the driver's seat.

You probably should focus on the metric that actually matters...

You mentioned "lousy ROAS", which is great—you're already thinking about profitability. But let's take it a step further. ROAS (Return on Ad Spend) is a great indicator, but it can be misleading. A 5x ROAS sounds amazing, but if it's on a £10 product, you've only made £40 in profit before other costs. A 2.5x ROAS on a £200 product is far more valuable. The real question you should be asking isn't "What's my ROAS?", but "How much can I afford to pay to acquire a customer and still be very profitable?"

The answer to that lies in your Customer Lifetime Value (LTV). This is the total profit you expect to make from a single customer over the entire course of their relationship with your business. Once you know this number, everything changes.

Let's break it down. You need three pieces of information:

1. Average Revenue Per Account (ARPA): Or in your case, Average Order Value (AOV). What's the average amount a customer spends in one purchase?
2. Gross Margin %: After the cost of goods sold, what's your profit margin on each sale? Be honest here.
3. Monthly Churn Rate: This is a bit trickier for ecommerce. A better way to think of it is repeat purchase rate. How many customers come back to buy again within, say, 12 months? If 30% of customers buy again, your retention rate is 30% and your churn is effectively 70% for that period. For simplicity, let's stick to the classic churn model which is more for subscription, but the principle is the same. Let's imagine what percentage of your customer base you 'lose' each month (i.e., they don't come back). A low number is good.

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

Let's say your AOV is £80, your gross margin is 60%, and your monthly churn is 5% (meaning you retain 95% of your customers month on month, which is very good).
LTV = (£80 * 0.60) / 0.05
LTV = £48 / 0.05 = £960

Suddenly, your perspective shifts. Each customer is worth £960 in profit to you over their lifetime. A common rule of thumb is to maintain a 3:1 LTV to Customer Acquisition Cost (CAC) ratio. This means you can afford to spend up to £320 (£960 / 3) to acquire a single customer and still have a very healthy, scalable business. Your target CPA isn't £10 or £20 anymore; it's £320. A ROAS that seemed 'lousy' before might actually be incredibly profitable when viewed through the lens of LTV.

Use the calculator below to get a feel for your own numbers. This isn't just an academic exercise; it's the foundation of a truly scalable advertising strategy. It tells you how much firepower you really have.

You'll need a better way to think about audiences...

Now that we've sorted the structure and the metrics, let's talk about who you're actually showing these ads to. You mentioned you're running broad targeting, which is fine for scaling *after* you've fed the pixel a lot of data. But for testing and finding new pockets of customers, you need to be more methodical.

I usually structure audiences based on the marketing funnel: Top of Funnel (ToFu), Middle of Funnel (MoFu), and Bottom of Funnel (BoFu). This ensures you're not just screaming into the void but are talking to people with different levels of awareness about your brand.

Top of Funnel (ToFu) - Cold Audiences:
These are people who have never heard of you. Your goal here is to introduce them to your brand and products. This is where you test your core assumptions.

• Detailed Targeting: Don't just target "fashion". Get specific. Target followers of complementary (but not competitor) brands, magazines your ideal customer reads, influencers they follow. Think about their lifestyle. For example, in one campaign for a women's apparel brand, we achieved a 691% return. Success in cases like that often comes from understanding the customer's lifestyle and targeting their broader interests, not just obvious ones related to clothing. It's about finding the psychographics, not just demographics.
• Lookalike Audiences: Once you have enough data (at least 100 purchases, but ideally 1,000+), this is your goldmine. Start with a 1% Lookalike of your Purchasers list. This is the highest quality audience you can build. Then you can test lookalikes of people who Added to Cart, or even all Website Visitors.

Middle of Funnel (MoFu) - Warm Audiences:
These people know you exist but haven't bought yet. They've visited your site, watched a video, or engaged with your Instagram page. The goal is to bring them back and get them to take the next step.

• Website Visitors: Retarget everyone who has visited your site in the last 30-90 days (but EXCLUDE purchasers). Show them testimonials, user-generated content, or highlight a specific benefit they might have missed.
• Video Viewers: Anyone who watched 50% or more of your video ads is clearly interested. Retarget them with a different ad, maybe a carousel showing the product in different settings.
• Social Engagers: People who have liked, commented, or shared your posts on Instagram or Facebook. They're warm leads; give them a reason to click through to the site.

Bottom of Funnel (BoFu) - Hot Audiences:
These people are on the verge of buying. They've added a product to their cart or started the checkout process. Your job is to get them over the finish line.

• Add to Cart (Last 7-14 Days): This is your most important retargeting audience. They were *this close*. Remind them what they left behind. Maybe offer a small incentive like free shipping if you can afford it, but often just a reminder is enough.
• Initiate Checkout (Last 7 Days): Even hotter than Add to Cart. Something spooked them at the last second. Show them an ad that reinforces trust – mention your return policy, customer reviews, or secure checkout.

You should have separate campaigns or ad sets targeting each part of this funnel. Your messaging and creative will be different for each. You don't talk to a complete stranger (ToFu) the same way you talk to someone with an item sitting in their online shopping basket (BoFu). By structuring your audiences this way, you create a complete journey for your customers, guiding them from awareness to purchase and beyond.

This is the main advice I have for you:

I know this is a lot to take in, so I've broken down the immediate actions I'd recomend you take. This isn't about quick fixes; it's about building a proper, sustainable system for growing your business with paid ads.

It's time to stop gambling and start engineering...

Look, what you've experienced is the default for most people who try to run Meta ads themselves. You follow the prompts, set up what seems like a logical campaign, and then get taken for a ride by an algorithm that doesn't share your goals. It feels like you're pulling a slot machine lever and hoping for a jackpot.

The shift needs to be from gambler to engineer. An engineer doesn't hope; they build a system. They understand the inputs and outputs. They test variables in a controlled way. They measure what matters. They build a machine that produces a predictable result, and then they pour fuel on it.

Everything I've outlined above—the ABO/CBO structure, the focus on LTV, the audience funnel—is about building that machine. It takes more work upfront than just throwing a few ads into a broad CBO campaign, but the result is control, predictability, and most importantly, profitable scale.

This is precisely the kind of work we do for our clients every day. We come in, diagnose the core issue (which is often very similar to yours), and then build a robust, data-driven system that turns their ad spend from an expense into an investment. For instance, by implementing these exact principles, we helped a software client reduce their cost per acquisition from £100 down to £7, and worked with an eCommerce subscription box brand to achieve a 1000% return on their ad spend.

If you're feeling a bit overwhelmed by all this and would rather have an expert build and manage this machine for you, I'd be happy to chat. We offer a free, no-obligation initial consultation where we can go through your ad account together and lay out a specific, actionable plan for you.

Either way, I hope this detailed breakdown has been genuinely helpful and gives you a clear path forward.

Regards,

Team @ Lukas Holschuh