Solved: Meta Ads ROAS Tanked After Scaling Budget

Hi there,

Thanks for reaching out! It’s a common problem you're facing, one I've seen countless times. You find a winning formula, you scale up, and then suddenly the whole thing falls off a cliff. It's incredibly frustrating, and it feels like the algorithm has turned against you. But it's usually not the algorithm being random; it’s a sign that the foundations of the campaign weren't quite ready for that level of pressure.

Happy to give you some initial thoughts on what's likely going on and how you might fix it. The short answer is that you've probably hit a scaling wall because you've exhausted your most profitable audience segment, and simply throwing more money at the same setup won't work. We need to get a lot more strategic about your audience, your creative, and your overall campaign structure.

The Scaling Myth: Why More Money Isn't the Answer

First things first, let's bust a myth. Scaling an ad campaign isn't just about cranking up the daily budget. That's not scaling; that's just spending more. When you do that, you're telling Meta, "find me more people like the ones who've been converting, but do it faster and in greater numbers."

What happens then? The algorithm's first port of call is to expand its reach within your targeted audience. It's already shown your ads to the most likely buyers – the people who are a perfect fit, the low-hanging fruit. They were responsible for your juicy ROAS of 4-5. But when you demand more volume, Meta has to start showing your ads to the next tier of people. These are the "maybe" people. They're a bit less interested, a bit more sceptical, a bit slower to buy. And so, your cost per purchase goes up, and your ROAS goes down. This isn't the algorithm failing; it's the algorithm doing exactly what you asked it to do with the audience you gave it. It's a sign of audience saturation.

You saw this happen textbook-style. At €1k/day you were likely at the edge of that prime audience. Pushing to €2k/day forced the algorithm deep into the expensive, lower-quality pockets of your audience, and the ROAS tanked. When you pulled back to €1k, the algorithm didn't just snap back. It has now learned from data that includes these less-receptive users, and the performance at that level is now 'polluted' by the recent push. It's not permanent, but it shows that the old 'sweet spot' is gone. You've burnt through the best bit of that audience.

We'll need to look at your audience strategy... or lack thereof

This brings me to the absolute core of the problem: your audience. You've got a winning creative, which is great, but a great creative shown to the wrong people is just a waste of money. To scale properly, you need a portfolio of audiences that you can tap into, not just one. You need to structure your campaigns to talk to people at different stages of awareness.

I usually structure this as a funnel: Top of Funnel (ToFu), Middle of Funnel (MoFu), and Bottom of Funnel (BoFu).

BoFu (Bottom of Funnel): These are your hottest prospects. People who have added a product to their cart, initiated checkout, or visited your checkout page but didn't buy. They are screaming "I want to buy this!". You should be hitting them with retargeting ads that remind them what they left behind, maybe offer a small incentive, and create urgency. Your budget here doesn't need to be huge, but the ROAS should be spectacular.

MoFu (Middle of Funnel): These are people who've shown some interest. They've visited your website, watched a chunk of your video ads, or engaged with your social media page. They know who you are but aren't ready to buy yet. You retarget them with ads that build trust and showcase different benefits of your products. Case studies, reviews, user-generated content (UGC) work brilliantly here.

ToFu (Top of Funnel): This is your cold traffic. People who've never heard of you. This is where the bulk of your scaling budget will go. This is also where things get tricky, and it's where your current campaign is likely failing. You can't just have one broad audience here. You need to be testing constantly.

Here's the priority I'd reccomend for testing your ToFu audiences:

1. Detailed Targeting: Start here. Get super specific. Don't just target "Fashion". Target followers of specific boutique brands that match your style, readers of niche fashion blogs, people interested in "sustainable clothing" AND "online shopping". Group related interests into themed ad sets and test them against each other.
2. Lookalike Audiences: Once you have enough data (at least 100 purchases, but honestly, you want more like 1,000 to get a really good quality audience), you need to build lookalike audiences. These are your secret weapon for scaling.
   • Start with a Lookalike of your previous customers. This is gold. Meta will go and find people who share characteristics with the people who have already given you money.
   • Then, create Lookalikes of people who initiated checkout.
   • Then, a Lookalike of people who added to cart.
   • Then, a Lookalike of your website visitors.
   Test these in order of priority. A lookalike of purchasers will almost always outperform a lookalike of website visitors. Test different percentages too (1%, 1-3%, 3-5%). Start narrow with 1% and broaden out as you need to scale.
3. Broad Targeting: This is what you only do when your pixel is really, really smart. Once you have thousands of purchases logged, you can sometimes run a campaign with very minimal targeting (e.g., just age, gender, location) and let the algorithm do the work. But you're not there yet. Doing this too early is just setting your money on fire.

You need to be constantly testing new audiences at the ToFu level to find new pockets of profitable customers. When one starts to fatigue, you swap it out for a new winner. Scaling isn't about one ad set; it's about managing a portfolio of them.

I'd say you need to rethink your creative strategy

You said, "The creative is still performing well — no signs of ad fatigue. Frequency is only 1.6". I'm going to challenge you on this. A frequency of 1.6 across the entire campaign lifetime is very low and tells me you've been reaching new people, which is consistent with your scaling. But ad fatigue isn't just about frequency. It's about creative saturation within a specific audience segment.

Your "winner" creative likely worked wonders on that initial, perfect audience. But the people you're trying to reach now, the more sceptical ones, might need a different message. What convinces an early adopter won't necessarily convince a laggard. You can't just keep showing the same ad to everyone and expect the same results.

Adding more creatives to the same ad set is a good idea, but it should be done with purpose. Don't just add variations of the same ad. You need to test fundamentally different angles.

• Test different hooks: If your current ad starts with the product, make a new one that starts with the problem it solves.
• Test different formats: You have 3 creatives. Are they all images? Test a video ad. Test a carousel ad. We've seen some of our SaaS clients get amazing results with simple user-generated content (UGC) style videos. For one client, a medical job matching platform, a strategic creative overhaul was a key part of how we reduced their Cost Per User Acquisition from £100 down to just £7.
• Test different messages: Create ads for different funnel stages. A cold audience (ToFu) might need an ad that educates them on the problem. A warm audience (MoFu) that visited your site might need an ad with customer testimonials to build trust. A hot audience (BoFu) that abandoned their cart needs an ad that says "Did you forget something?".

Your current approach of one winning ad set is a single point of failure. A succesful scaling strategy relies on a robust creative testing system. You should always have a small portion of your budget dedicated to finding your next "winner" creative, because I can promise you, the current one's performance will eventually decline. It's not a matter of if, but when.

You probably should calculate your real numbers

Here's the biggest mindset shift you need to make to scale succesfully. Stop obsessing over daily ROAS. Daily ROAS is a vanity metric. It fluctuates wildly and causes people to make panicky, short-sighted decisions (like cutting a budget on a bad day when it might have recovered).

The only two numbers that truly matter are:
1. Customer Acquisition Cost (CAC): How much it costs you, on average, to get one new customer.
2. Lifetime Value (LTV): How much profit a customer will generate for you over their entire relationship with your business.

Once you know your LTV, you know how much you can afford to spend on your CAC. Let's do some quick maths. Say your average order value is €80, and your gross margin is 60%. A customer buys from you, on average, 3 times a year and stays a customer for 2 years.

LTV = (Average Order Value) * (Purchases per Year) * (Years as Customer) * (Gross Margin)
LTV = (€80) * (3) * (2) * (0.60) = €288

So, each customer is worth €288 in profit to you. A healthy business model aims for an LTV:CAC ratio of at least 3:1. This means you can afford to spend up to €96 (€288 / 3) to acquire a single customer and still have a very profitable business.

Suddenly, a day where your ROAS is 1.8 doesn't look so bad, does it? If your Average Order Value is €80, a ROAS of 1.8 means your cost per purchase (your CAC) is about €44 (€80 / 1.8). That's well below your allowable CAC of €96! You're still making money hand over fist, but your obsession with a high daily ROAS is blinding you to it.

Knowing this number frees you. It allows you to invest in acquiring customers at what might seem like a "loss" on the first purchase, knowing you'll make it back over time. This is the key to out-spending and out-manoeuvring your competition. Use the calculator below to get a feel for your own numbers.

You'll need a new action plan

So, where do you go from here? Cutting the budget back to €500/day isn't a bad temporary move to stop the bleeding and give you breathing room. But it's not a long-term strategy. Here is what I would do.

Phase 1: Stabilise & Analyse (Next 7 days)

1. Scale back to €500/day. Let the algorithm resettle and gather some baseline data at this more stable level. Don't touch it for a few days.
2. Do a deep dive on your funnel. Where are people dropping off? Low Click-Through Rate (CTR)? Your ad isn't grabbing attention. High CTR but low Add-to-Carts? Your landing page or product page has a problem. Lots of Add-to-Carts but few purchases? Your checkout process is broken or your shipping costs are a surprise. Fix the leaks in your bucket before you try to pour more water in.
3. Calculate your LTV. Get a real, data-backed number for your allowable CAC. This will be your north star for every decision you make going forward.

Phase 2: Rebuild & Test (Weeks 2-4)

1. Build out a proper funnel structure. Create separate campaigns for ToFu, MoFu, and BoFu.
   • BoFu Campaign: Retarget 'Add to Carts' and 'Initiated Checkouts' from the last 14 days. Use Dynamic Product Ads. Allocate 10-15% of your budget here.
   • MoFu Campaign: Retarget website visitors and video viewers from the last 30 days. Show them testimonials and different product use-cases. Allocate 15-20% of your budget.
   • ToFu Campaign: This is where you test. Create 3-4 new ad sets. One with your original "winning" audience. The others should be new, highly specific interest-based audiences. And if you have the data, your first customer lookalike audience. Allocate the remaining 65-75% of your budget here.
2. Launch a creative testing campaign. Use a small, separate budget (€50/day) to test new creative angles constantly. Find new winners to feed into your main ToFu campaign. This is your R&D department.

I've detailed my main recommendations for you below in a table to make it clearer.

This is obviously a lot to take in, and implementing it correctly takes time and experience. The difference between a professional setup and a DIY approach can be huge. I remember one client we worked with, an eCommerce business selling women's apparel, who saw a 691% return on ad spend after we restructured their campaigns using these exact principles. Another, selling subscription boxes, hit a 1000% ROAS. It's not magic; it's just a rigorous, data-driven process that most people don't have the time to learn and execute.

Scaling ads is where most businesses get stuck and end up burning a lot of cash. Getting expert help can often be the difference between hitting a plateau and breaking through to the next level of growth.

If you'd like to chat through this in more detail and have us take a proper look at your ad account, we offer a free 20-minute strategy session. We can audit your campaigns together and give you a more tailored plan of action. No obligation, of course, but it might be exactly what you need to get things back on track.

Hope this helps give you some clarity.

Regards,

Team @ Lukas Holschuh