Solved: Facebook Ads Performance Suddenly Dropped Off

Hi there,

Thanks for reaching out! What you're describing is an incredibly common, and frustrating, part of the learning curve with Meta ads. It can feel like the platform is actively working against you, one day giving you a glimpse of success and the next snatching it away. I'm happy to give you some initial thoughts and guidance based on my experience running these sorts of campaigns. The short answer is that the problem isn't that day two and three were bad, the problem is that you've been led to believe day one was real. It almost certainly wasn't, and making decisions based on it is the fastest way to burn through your budget.

Let's break down what's likely going on and how you can build a more stable, predictable system for growth.

We'll need to look at why that first day was probably just a fluke...

This is the first and most difficult pill to swallow. A $50 spend and 6 sales for a 4 ROAS is not a "winner". It's a statistical anomaly. It's a blip. It's a lucky punch. When you give Meta's algorithm a brand new ad set with a tiny budget, it goes into an exploration phase. It doesn't know who your customer is yet, so it starts showing your ad to small pockets of people within your targeting parameters to see what sticks. On day one, it just so happened to stumble upon a tiny, cheap, and unusually responsive group of people who were ready to buy at that exact moment. You got lucky.

The problem is, you interpreted this luck as a reliable signal. You thought you'd struck gold, so you did what seemed logical: you tried to dig faster by doubling the budget. But by changing the budget from $50 to $100, you sent a shockwave through the system. Meta's algorithm, which was happily fishing in its tiny, productive pond, was suddenly told "Right, that's not enough, I need you to go find me twice as many people, right now." This forced it out of that lucky pocket and into the wider, more expensive, and less interested ocean of your total target audience. It had to start exploring all over again, but this time with more pressure and a higher spend rate. This is why your CPM skyrocketed and your metrics fell off a cliff. Any "learning" from day one was instantly wiped out.

Making any judgements, good or bad, based on 24-48 hours of data is one of the biggest mistakes I see new advertisers make. It leads to this exact emotional rollercoaster and a strategy built on guesswork and reaction. We need to get you off that ride and into a system based on process and patience. I've worked on campaigns for eCommerce brands that have generated huge returns, like a 691% return for a women's apparel brand or a 1000% ROAS for a subscription box, but none of those results came from a single good day. They came from weeks of methodical testing and data analysis, weathering the bad days to understand what truly works in the long run. Some days the ads just don't perform, that's part of the game. Your job isn't to react to every dip, but to build a structure that is resilient to them.

I'd say you need a proper testing framework, not just reacting to daily swings...

So, how do we do that? You need to stop thinking in terms of single campaigns and start thinking in terms of a funnel. Right now, you're just throwing one hook in the water and getting frustrated when you don't catch a prize fish every time. A proper structure has different campaigns for different stages of the customer journey. For a new eCom store, it's pretty simple to start.

I'd recomend a structure that looks something like this:

1. Top of Funnel (ToFu) - Prospecting: This is where you are now. The goal here is to find new customers, people who have never heard of you. Your entire focus in this campaign should be on testing different cold audiences to see which one responds best. You shouldn't just have one ad set, you should have several, each testing a different hypothesis about your customer.

2. Middle/Bottom of Funnel (MoFu/BoFu) - Retargeting: This is the campaign you're missing, and it's probably where most of your profit will come from. This campaign's only job is to show ads to people who have *already* visited your website or engaged with your ads from the ToFu campaign but didn't buy. Most people don't buy on the first visit. They need to see your brand a few more times. A retargeting compaign makes sure that happens.

When we take on a new client, we audit their account and almost always find they are not prioritising their audiences correctly. They're either spending all their money at the top of the funnel or they're running messy retargeting that doesn't make sense. You need to be systematic. The further down the funnel someone is, the more valuable they are, and the more likely they are to convert.

Here’s a rough priority list for audiences that we would typically test for an eCommerce client. You won't have the data for most of these yet, but this is what you're building towards:

Your current strategy is just playing in that last box. You need to build the infrastructure to capture everyone else. And inside your ToFu campaign, you should launch with 3-4 different ad sets, each with a distinct audience, and let them run for at least 5-7 days without touching them. Only then can you start to see which audience is *actually* a winner, not just a one-day fluke. You'll definately see some ad sets fail, and that's fine. That's data. You're paying to learn which audiences to turn off so you can put more money behind the ones that work consistently.

You probably should rethink your entire approach to targeting...

Let's talk about those ToFu audiences. You said your creatives were "solid enough" which we'll get to, but your targeting is the other half of the equation. A common mistake is picking interests that are too broad or generic.

You need to forget the idea of a demographic. "Women aged 25-45 who like online shopping" is not a target audience. It's a meaningless collection of millions of people. You need to define your Ideal Customer Profile (ICP) by their *pain*. What is the specific, urgent, nagging problem that your product solves? Who is the person feeling that pain most acutely?

Once you understand their nightmare, you can find them. They aren't just "people who like jewelry". They're people who follow specific independent designers on Instagram, who read niche fashion blogs, who buy from competitor brands, who have shown interest in "sustainable fashion" or "artisan goods". You have to go deeper.

For example, if you were selling high-quality, durable outdoor gear, targeting an interest like "Camping" is a terrible idea. You'll get millions of people who went camping once five years ago. But targeting interests like "Patagonia", "The North Face", "Backpacking Magazine", and layering that with behaviours like "Engaged Shoppers" gets you much, much closer to someone who is actually in the market for what you sell. I remember one campaign for an outdoor equipment brand where this exact shift in targeting thinking was what allowed us to generate over 18,000 website visitors and build a massive retargeting pool that drove sales for months.

Your job in the prospecting phase is to create 3-4 of these audience "hypotheses" and test them against each other. For instance:

• -> Ad Set 1 (Competitor Fans): Target interests of 3-5 direct competitor brands.
• -> Ad Set 2 (Publication Readers): Target interests of 3-5 niche magazines or blogs your ICP reads.
• -> Ad Set 3 (Adjacent Interests): Target interests of complementary products or activities your ICP enjoys.

By splitting them up, you let Meta's algorithm find the cheapest conversions within each distinct group. After a week, you'll have real data on which customer profile is most profitable, and you can scale from there.

You'll need to understand the numbers that drive growth...

This brings me to the next point. You are obsessing over the wrong metrics. CPM, CTR, daily ROAS... on day three, these are vanity metrics. They are distractions. The only math that truly matters for scaling a business is the relationship between your Customer Lifetime Value (LTV) and your Customer Acquisition Cost (CAC).

The question isn't "how cheap can I get a sale?". It's "how much can I *afford* to pay for a sale and still be wildly profitable?". Understanding this will free you from the stress of daily performance swings.

Let's do a quick, back-of-the-napkin calculation. I don't know your exact numbers, so let's make some up for your product:

• -> Average Order Value (AOV): Let's say it's £40.
• -> Gross Margin: After product costs, what's left? Let's say 70%. So your Gross Profit per order is £28.
• -> Repeat Purchase Rate: This is the big one. How many customers buy again? Let's be conservative and say 1 in 5 customers (20%) buy one more time within a year.

So, your average customer is worth: (1 * £28) + (0.20 * £28) = £28 + £5.60 = £33.60. This is a very basic LTV.

A healthy business model aims for a 3:1 LTV to CAC ratio. This means you can afford to spend up to £33.60 / 3 = £11.20 to acquire a single customer.

Suddenly, the picture looks different, doesn't it? On day one, with a $50 spend (£40) and 6 sales, your CAC was £40 / 6 = £6.67. This is well below your affordable CAC of £11.20. It was extremely profitable.

On day two and three, maybe you spent £80 and got 2 sales. Your CAC shot up to £40. This is way above your affordable threshold. But now you see the problem isn't just "the ads stopped working". The problem is your CAC is volatile because your strategy is based on luck. The goal of a proper advertising system is to find audiences and creatives that can *consistently* deliver a CAC below that £11.20 mark at scale.

To give you some real-world context, here are some typical cost-per-result ranges we see. For an eCommerce store selling to developed countries, getting a purchase is not cheap.

As you can see, your day one CPA of £6.67 was exceptionally low. Your performance on days two and three, while not profitable, is much more indicative of the challenge of advertising in competitive markets. Your goal is to build a machine that lives in the profitable end of that £10-£75 spectrum, not to hope for another £6 miracle.

We'll need to look at your offer and creative, because that's your biggest lever...

Finally, let's talk about the one thing you have complete control over: your ads and your offer. You said yourself, "I wasn’t super in love with the creatives". This is a massive red flag. If you, the person who knows the product best, are not even excited by the ad, why would a complete stranger scrolling through their feed give it a second of their time?

"Solid enough" is the enemy of performance. You need creative that stops the scroll and a message that connects instantly. You need to stop selling a product and start selling a transformation. You can use simple copywriting frameworks for this.

Problem-Agitate-Solve:

Problem: Tired of generic, mass-produced accessories?
Agitate: Everywhere you look, it's the same old stuff. It's impossible to show your unique style.
Solve: Discover our handcrafted [Your Product Name]. Each piece is unique, just like you. Shop the collection.

Before-After-Bridge:

Before: Your outfit feels incomplete, missing that one final touch.
After: Imagine turning heads with a stunning, unique piece that tells a story.
Bridge: Our [Your Product Name] is the bridge. Handcrafted to make your style unforgettable.

These simple structures force you to focus on the customer's emotional state, not just the product's features. You need to be testing these different angles relentlessly. Every week, you should be putting new creatives into your prospecting campaign to try and beat your current best-performer. Test images vs. videos. Test user-generated style content. Test different headlines. This is how you lower your CAC over time.

And remember, the ad is only half the battle. When people click, they land on your product page. Does it look trustworthy? Are there customer reviews? Is the photography professional? Is the product description compelling? The eCommerce critique I often give is that many new stores feel untrustworthy, which kills conversion rates. Meta's algorithm is smart; if it sees people clicking your ad and immediately bouncing from your website, it will correctly assume your user experience is poor and charge you more to show your ads. Improving your store's conversion rate is just as important as improving your ads.

I've detailed my main recommendations for you below:

This is a lot to take in, I know. It's a fundamental shift from gambling to engineering. It's easy to burn through a lot of money very quickly trying to figure all this out on your own through trial and error. The principles are simple, but the execution requires discipline and experience.

This is exactly the kind of situation we help our clients with every day. We build these systems for them, manage the testing process, and help them scale predictably without the day-to-day anxiety. If you'd like to chat further, we offer a free, no-obligation strategy session where we can have a proper look at your account and website together and build a clear roadmap to get you past these initial hurdles and onto a path of consistent growth.

Regards,

Team @ Lukas Holschuh