Solved: Facebook Ads Campaign Performance Drop

Hi there,

Thanks for reaching out! Happy to give you some initial thoughts and guidance on your campaign troubles. What you're describing, a campaign that works brilliantly and then suddenly falls off a cliff, is frustratingly common with Meta ads. It's something we see all the time. One minute you're celebrating a 7x ROAS, the next you're pulling your hair out wondering where all the sales went.

It's always a bit tricky to give super specific advice without actually looking inside the ad account itself, as there are so many moving parts. But based on what you've said, I can definitely walk you through how I'd approach this situation and what the most likely culprits are. This should give you a solid framework for figuring out what to do next.

We'll need to look at why the campaign stopped performing...

First off, let's talk about the 'why'. When a campaign dies overnight like yours did, there's usually a few potential reasons. It's almost never just one thing, but a combination of factors. The most common one is simple ad fatigue. Your audience, even a really good one, has just seen your ads too many times. They've either bought, or they've decided they're not interested, and now they just scroll past. It's like a joke that stops being funny the tenth time you hear it.

You mentioned you upped the budget twice during February. That's great when things are working, but it also accelerates this process. When you increase the spend, Meta's algorithm works harder to find people to show your ads to. It burns through the 'low-hanging fruit' – the people most likely to buy – much, much faster. Once that core group is exhausted, the algorithm has to branch out to a slightly less interested audience to spend the new, higher budget. This almost always leads to a drop in performance and higher costs per purchase. So the budget increase, while profitable at first, likely sowed the seeds for this sudden drop.

Then there's the Meta algorithm itself. Tbh, it's a bit of a black box. Sometimes performance just drops for no obvious reason. There could be an unannounced tweak to the system, or maybe a load of new advertisers entered the auction for your audience, driving up your costs. I remember one campaign we ran for a client in the gambling niche, doing prize draws, that was giving us a steady 618% return for ages. Then one week, it just dipped. We had to go back to the drawing board. It's just the nature of the beast, you can't get too attached to a single winning campaign because it will almost certainly stop working at some point.

Two days is a very short window to judge, but a drop from 7x ROAS to unprofitable is a massive red flag. You're right to be concerned and not just let it run, hoping it'll fix itself. It rarely does. Waiting just means you're burning money that could be better spent on testing a new approach.

I'd say you should diagnose the drop-off...

Before you just turn things off or randomly start changing things, it's good to play detective for a bit. You need to look at the metrics in Ads Manager to see where the problem is. This tells you what to fix. Look at the last two days where it was unprofitable and compare them to a period in February when it was working great.

Here’s what you should be looking for:

-> Check your Ad-Level Metrics: The first place to look is at the ads themselves.

• Frequency: Is this number creeping up? If an average user has seen your ad 4, 5, 6+ times in a short period, you've almost certainly hit ad fatigue.
• Click-Through Rate (CTR): Has this plummeted? If people aren't clicking anymore, it's a huge sign the creative is stale. The image and the headline aren't grabbing them like they used to.
• Cost Per Click (CPC) / Cost Per Mille (CPM): Have these shot up? If your CPC or CPM (the cost for 1,000 impressions) has spiked, it means you're paying more to reach people. This could be because of increased competition in the auction, or because the algorithm is having to work harder (and spend more) to find people in your audience who will engage.

-> Check your Funnel Metrics: It’s not always the ad's fault. The problem might be happening after the click.

• Landing Page Views vs Clicks: Are you still getting a similar number of landing page views for every click? If you have lots of clicks but a big drop in landing page views, it could be a technical issue like your website loading too slowly.
• Adds to Cart: Are people still adding your product to their cart after visiting the product page? If you're getting lots of visitors but no one is adding to cart, something is wrong on the product page. Maybe the price is an issue, or the product photos or description aren't compelling enough for this new slice of the audience.
• Initiated Checkouts & Purchases: This is the big one. If you're getting plenty of adds to cart but very few people are starting or completing the checkout, there could be a problem in your cart or checkout process. Maybe a discount code isn't working, or there's a surprise shipping cost that's scaring people off. It's worth going through the checkout process yourself to make sure everything is smooth.

Looking at these metrics tells you where the leak in your funnel is. A drop in CTR points to creative fatigue. A drop in conversion rate on the website points to an issue with your offer, your page, or the quality of the traffic Meta is now sending you.

Here's a simple way to think about it:

By figuring out the 'what', you can make a much more informed decision about 'what next'.

You probably should restructure your campaigns...

So, to answer your main question: "Do I turn off the campaign? Or do I add new ads with new creatives inside this same campaign?". My answer is a bit of both, but with a specific process.

Do not just add new ads to the failing campaign. When an ad set has been running for a while, the algorithm has learned a lot about it. If it's now performing poorly, editing it or adding new ads can confuse the system and you'll be trying to fix something that's already broken. It's rarely effective.

Instead, here's the process I'd follow:

1. Pause the Underperforming Campaign: First thing's first, stop the bleeding. Turn the campaign off so you're not wasting any more money.

2. Duplicate the Campaign: Next, duplicate the entire campaign. This creates a fresh copy. By duplicating instead of editing, you give the algorithm a clean slate to work with. You're keeping the basic structure that once worked, but you're forcing it to enter a new 'learning phase' with the changes you're about to make.

3. Introduce New Variables: In this new, duplicated campaign, you need to systematically test new things. Don't change everything at once. Change one major thing at a time.

• Test New Creatives: Your old ads are probably burned out. Create a few new ones. If you were using images, try a simple video. If you were using polished studio shots, try something that looks more like user-generated content (UGC). We had a SaaS client selling to other businesses who saw amazing results just by switching to simple, authentic-looking UGC videos. The same principle applies to eCommerce. Show real people using or enjoying your product. Test different headlines and different calls to action.
• Test New Audiences: The audience that brought you a 7x ROAS is likely saturated. It's time to find new people. This is where a proper funnel structure comes in. You need to be thinking about your audiences in seperate stages.

For eCommerce, the best performing accounts I've worked on all have a clear seperation between prospecting for new customers (Top of Funnel - ToFu) and retargeting people who have already shown interest (Middle/Bottom of Funnel - MoFu/BoFu).

Your old campaign was probably doing a bit of everything. Now you can get more strategic. Here’s a list of audiences I would usually prioritize, from broadest to most specific. Since you were getting sales, you should have data to build some of the most powerful audiences on this list.

You should start by building out a new prospecting (ToFu) campaign and a seperate retargeting (BoFu) campaign. Put most of your budget into prospecting to find new customers, and a smaller amount into retargeting to convert the high-intent ones who slip away.

You'll need a long-term testing framework...

The key takeaway here shouldn't be just how to fix this one broken campaign, but how to stop it from happening again. Or rather, how to have a system in place for when it inevitably does. Paid advertising isn't a 'set and forget' channel. Your amazing February results prove the potential, but the sudden drop proves the instability.

A robust advertising strategy is built on continuous testing. You should always have new creatives and new audiences in the pipeline. I remember one campaign where we managed to get a 1000% ROAS for a subscription box client, and it wasn't by finding one magic ad and running it forever. It was by building a system where we were constantly testing new creatives and audiences, identifying winners, scaling them up, and then finding the next winner before the current one died out. I also recall working for another eComm client selling cleaning products, where we got a 633% return by having a solid ToFu/BoFu structure in place, which meant we could weather the small dips and keep the sales consistent.

Your job isn't to find one perfect campaign, but to build a machine that consistently finds good enough campaigns. This means having seperate campaigns for your different funnel stages, as I outlined above. Inside your prospecting campaign, you can have multiple ad sets, each testing a different audience (e.g., one for an 'Add to Cart' lookalike, one for an interest group). Inside each ad set, you should be testing 2-3 different creatives.

After a few days, you look at the data. Turn off the ads and audiences that aren't working. Give a bit more budget to the ones that are. And every week or two, you should be adding a new creative or a new audience into the mix to test. This way, you're never reliant on one single ad set. When one starts to fade (and it will), you already have data on what might work next.

This is the main advice I have for you:

I know that's a lot to take in. To make it simpler, I've detailed my main recommendations for you in a table below. This is the step-by-step plan I would implement if I were in your shoes.

This might all seem like a lot of work, and honestly, it is. Managing paid ad campaigns effectively is a full-time job that requires constant analysis, testing, and strategic thinking. You've already proven you have a product that people want and that ads can work for you – that 7x ROAS is the proof. The challenge now is building a more professional, resilient system to make those results more consistent and scalable.

This is where getting some expert help can make a huge difference. An experienced eye can spot opportunities and problems much faster and implement the kind of structures I've described above to take the guesswork out of it for you. It lets you focus on running your business, knowing the advertising is being handled strategically.

If you'd like to go over your account in more detail and get a proper hands-on review, we offer a free initial consultation where we can look at things together on a call. We could walk through your campaigns and I could give you some more specific pointers. No strings attached, of course. Just let me know if that's something you'd be interested in.

Hope this helps give you a clear path forward!

Regards,

Team @ Lukas Holschuh