Solved: Scaling an ABO Campaign After 2.10 ROAS

Hi there,

Thanks for reaching out!

Happy to give you some initial thoughts on your campaign situation. It's a classic spot to be in – you've got a bit of a positive signal with that 2.10 ROAS, and the temptation to hit the accelerator is huge. I see this all the time. But the difference between burning your budget and building a truly scalable campaign often comes down to what you do right now.

The short answer is you're asking the right questions, but probably a little too soon. Let's unpack why that is and what a more robust plan looks like.

You probably should... hold your horses for a bit longer

I know it's the last thing you want to hear when you see some green on the dashboard, but the most profitable move you can make right now is to do nothing. Seriously. Two and a half days is nowhere near enough data to make an informed decision. Here's why:

The Learning Phase: Meta's algorithm is smart, but it's not a magician. When you launch a new ad set, it enters what's called the "learning phase." During this time, it's actively trying to figure out who within your target audience is most likely to take the action you want (in this case, purchase). It needs about 50 conversions per ad set within a 7-day period to exit this phase. Until it does, your results will be all over the place. Your cost per purchase might be £5 on Monday, £50 on Tuesday, and £15 on Wednesday. Acting on any single day's data is like trying to navigate a ship in a storm by looking at a single wave. You need to wait for the waters to calm down a bit so you can see the actual tide.

Performance Fluctuates: Think about your own buying habits. Are you as likely to buy something on a Tuesday morning as you are on a Friday evening? Probably not. Performance naturally ebbs and flows throughout the week. You've only captured a tiny snapshot of user behaviour. A 2.1 ROAS on a Monday and Tuesday might turn into a 0.8 ROAS by the weekend, or it could shoot up to 4.0. You just don't know yet. You need a full 7 days of data, at a minimum, to get a true weekly average that smooths out these daily peaks and troughs.

Statistical Significance: Let's say your average cost per purchase is around $30. With a $70 daily budget spread across three ad sets, each ad set is spending about $23 a day. After 2.5 days, that's roughly $58 spent per ad set. That might mean you've only got one, maybe two purchases in that winning ad set. Is that enough to declare it a winner and bet the farm on it? Absolutely not. It could just be luck. You need a higher volume of conversions before you can say with any confidence that one ad set is genuinely better than another.

I remember one campaign for a women's apparel client where, after three days, one ad set had a nearly 7x return and another was barely breaking even. The temptation was to kill the loser and scale the winner. We advised them to wait. By day seven, the "winning" ad set had dropped to a 2x return, while the "loser" had found its groove and was consistently delivering a 4.5x return. Patience literally made them thousands of pounds. This is a common situation; you're reacting to noise, not a signal.

We'll need to look at... why just duplicating ad sets is a trap

Okay, let's fast forward. It's been a week, and that one ad set is still a clear winner. Now's the time to duplicate it and double the spend, right? Not so fast. This tactic, often preached in online courses, is one of the most misunderstood and misused scaling methods out there. It can work, but it can also be a spectacular way to ruin a good thing.

Think about what you're doing. You have Ad Set A targeting, let's say, women aged 25-40 who are interested in "Sustainable Fashion". It's working well. So you duplicate it, creating Ad Set B, which targets the *exact same* group of people. Now you have two of your own ad sets bidding for the attention of the same audience in the same auction. This is called audience overlap.

When this happens, you're essentially bidding against yourself. Meta's system is designed to prevent your own ad sets from directly competing too much, but it's not perfect. The result is often that your CPMs (the cost to reach 1,000 people) start to creep up across both ad sets. Your frequency might also rise faster, annoying the audience and leading to ad fatigue. You're trying to squeeze more juice from the same lemon, and often you just end up with a sour taste and higher costs.

The "duplicate and double" method is a form of vertical scaling – trying to spend more money on the same audience. A much safer and more sustainable approach, especially early on, is horizontal scaling – finding *new* winning audiences to spend money on. Instead of one audience at £40/day, your goal should be to find four different audiences that can each spend £10/day profitably. This diversifies your risk and gives you a much more stable foundation to scale from.

Here's a simple way to think about the decision process. It's not just "is it profitable?".

I'd say you... need to test creative properly, not just tweak it

Your idea to change the video by "swapping the person in the hook" is a good instinct. You're thinking about creative testing, which puts you ahead of many advertisers. But tbh, it's a bit like changing the colour of your shoelaces and expecting to run a faster marathon. It's a micro-iteration, and it's unlikely to give you the significant performance lift you're looking for.

To find a true winning creative, you need to test fundamentally different angles and formats. An 'angle' is the core message or the 'big idea' behind your ad. A 'format' is how you present that idea (video, image, carousel, etc.).

Let's imagine you're selling a high-end, organic face cream. Here are some different angles you could test:

• Angle 1: The Problem/Solution Angle. "Tired of dry, irritated skin? Our cream uses natural botanicals to restore your skin's moisture barrier." This speaks directly to a pain point. We often use the Problem-Agitate-Solve framework here. Problem: "Is winter weather leaving your skin feeling tight and flaky?" Agitate: "Most moisturisers are full of chemicals that only make it worse." Solve: "Our cream uses 100% organic ingredients to soothe and protect."
• Angle 2: The Social Proof Angle. "See why Vogue called our cream 'a miracle in a jar'." This leverages authority and trust. It could feature testimonials, user-generated content (UGC), or press mentions.
• Angle 3: The 'Before & After' Angle. A classic for a reason. Show the transformation. With the Before-After-Bridge framework, you paint a picture. Before: A picture of dull, unhappy skin. After: A picture of glowing, radiant skin. Bridge: "Our cream is the bridge that gets you there in just 14 days."
• Angle 4: The Ingredient Focus Angle. "Most creams use cheap fillers. We source our shea butter directly from a women's cooperative in Ghana. Here's why that matters for your skin..." This appeals to conscious consumers who care about quality and ethics.

You see how these are worlds apart from just changing the person in the first three seconds? Each one speaks to a different customer motivation. Once you have your angles, you can then test them across different formats. Maybe the Problem/Solution angle works best as a direct-to-camera video, while the Social Proof angle is perfect for a carousel ad showing off five different 5-star reviews.

I recall one campaign for a luxury brand launch where the goal was massive reach and impact. We didn't just run one video. We tested cinematic brand films against slick, fast-cut product showcases, and even simple, elegant static images. It was this combination of formats that allowed us to hit over 10 million views, because different people in the audience responded to different types of creative.

Here’s a sample testing matrix you could use. The goal is to fill this out to get true creative diversity.

You'll need... a structured framework for testing audiences

So if duplicating isn't the immediate answer, and you need to test creative properly, where does that leave you? It leaves you needing a proper structure. Most advertisers fail because they just throw random ad sets at the wall to see what sticks. Professional advertisers build a machine. That machine is the marketing funnel, broken down into campaigns for each stage.

Here’s how I would structure it for any eCommerce account, from a small startup to a multi-million pound brand. It's based on audience temperature: cold, warm, and hot.

1. Top of Funnel (ToFu) - Prospecting Campaign (Cold Audiences)

This is where you find new customers. These people have likely never heard of you before. Your $70/day ABO campaign is a ToFu campaign. The goal here is broad reach, but intelligent broad reach. Inside this campaign, you'd test different ad sets, each with a different cold audience.

• Interest-Based Audiences: This is your bread and butter to start. But you need to be smart. If you're selling coffee beans, don't just target "Coffee". That's millions of people, most of whom are happy with their Nescafe. Instead, target interests like "James Hoffmann", "Third Wave Coffee", "Aeropress", or followers of specific roaster pages. Get niche and relevant.
• Lookalike Audiences (LALs): This is where the magic happens once you have data. A Lookalike Audience is where Meta takes a "source" audience you provide and finds millions of other users who share similar characteristics. You should prioritise your source audiences based on value. The hierarchy looks like this:
  -> LAL of your Highest Value Customers (use a customer list)
  -> LAL of all Purchasers
  -> LAL of people who Initiated Checkout
  -> LAL of people who Added to Cart
  -> LAL of all Website Visitors

You start at the bottom of that list and work your way up as you gather more data. You need at least 100 people in a source audience to create a lookalike, but tbh, it works much better when you have 1,000+.

2. Middle of Funnel (MoFu) - Engagement Campaign (Warm Audiences)

These people have shown some interest. They know you exist, but they're not ready to buy. They've visited your website, watched one of your videos, or liked an Instagram post. The goal here is to build trust and stay top of mind. Your ad sets here would target:

• Website Visitors from the last 30 days (excluding anyone who has purchased).
• People who watched 50% or more of your video ads in the last 90 days.
• Instagram & Facebook Page Engagers from the last 90 days.

The ads you show them are different. Not "Buy Now!", but maybe a video explaining your brand story, a carousel of customer testimonials, or a link to a helpful blog post.

3. Bottom of Funnel (BoFu) - Retargeting Campaign (Hot Audiences)

This is where the money is. These people are on the verge of buying. They've added a product to their cart but got distracted, or they started the checkout process but didn't finish. These are your highest-intent audiences, and you should be hitting them with direct, compelling offers.

• Added to Cart in the last 7 days (excluding purchasers).
• Initiated Checkout in the last 7 days (excluding purchasers).
• Viewed Content (product pages) in the last 14 days (excluding purchasers and add to carts).

The ads here should be urgent and clear. Use Dynamic Product Ads (DPA) to show them the exact product they left behind. Maybe offer a small discount or free shipping to nudge them over the line. "Still thinking about it? Here's 10% off to help you decide."

We'll need to look at... understanding the numbers that actually matter

Finally, let's talk about that 2.10 ROAS. It's a good start, but ROAS is a lagging indicator. It tells you the result, but it doesn't tell you *why* you got that result. To really understand performance and know what to fix, you need to look further up the funnel at your diagnostic metrics.

• Click-Through Rate (CTR - Link): This tells you if your ad creative and copy are grabbing attention. A good CTR (usually 1% or higher) means people are interested enough in your ad to click. If this is low, your creative is the problem.
• Cost Per Click (CPC - Link): This tells you how much you're paying for that interest. A high CPC can mean your audience is very competitive, or that Meta thinks your ad isn't very relevant (low Quality Score). If this is high but your CTR is good, your audience is probably too expensive.
• Cost Per Add to Cart: This is a brilliant metric. It bridges the gap between the ad and your website. If you have a great CTR and a decent CPC, but very few people are adding to cart (or the cost is huge), then the problem isn't your ad. It's your landing page. The product description, the photos, the price, the trust signals – something there is putting people off.

A 2.10 ROAS might seem okay, but what if your business needs a 3.0 to be profitable after you account for product costs, shipping, and other overheads? And more importantly, what's the long-term value of that customer? A customer who buys once is good. A customer who buys five times over the next two years is gold. This is where understanding Customer Lifetime Value (LTV) becomes critical. It tells you how much you can *truly* afford to spend to acquire a customer.

Your current 2.10 ROAS might look average, but if those customers have a high LTV, you could actually be incredibly profitable in the long run. Suddenly, you can afford a higher cost per acquisition. This interactive calculator can help you reframe what you think of as "expensive".

I've detailed my main recommendations for you below:

So, to bring this all together, here's the exact plan I would follow if I were in your shoes. It's a shift from reactive tweaking to a proactive, structured process. This is how you build campaigns that last.

As you can see, there's a lot more to it than just duplicating an ad set. The most successful advertisers are patient, methodical, and operate with a clear system. They understand the numbers, they test intelligently, and they build structures that are designed for sustainable growth, not just quick, lucky wins.

This is just a high-level overview based on your situation. A proper deep-dive into your account, your analytics, and your business goals would likely uncover a dozen other opportunities for optimisation. This is where expertise really makes a difference – it's the ability to quickly diagnose the core issue and apply a proven framework to fix it, saving you weeks of wasted time and thousands in ad spend.

If you'd like to go through your account together on a call, we offer a completely free, no-obligation strategy session where we can do just that. It's often the fastest way to get clarity and an actionable plan.

Hope this helps!

Regards,

Team @ Lukas Holschuh