Solved: Custom Audience Ads Not Spending Budget

Hi there,

Thanks for reaching out about the issue with your custom audience ads. It's a really frustrating situation when a campaign that was working perfectly suddenly grinds to a halt, especially when you can't see an obvious reason why. I'm happy to give you some of my initial thoughts and guidance on what's likely going on and, more importantly, how to fix it.

From what you've described, this isn't a technical glitch or a bug. The slow spend is a symptom of a much more common problem that nearly every advertiser hits at some point: audience saturation and ad fatigue. Essentially, the Meta algorithm is telling you it can no longer find people within your warm audience who are likely to convert at a cost it deems acceptable. It's protecting your budget, but in the process, it's killing your campaign's delivery. We'll need to dig into why that is and how to give the algorithm fresh, fertile ground to work with.

We'll need to look at the real problem, not the symptom...

The first mistake most advertisers make in this situation is to start tinkering with bids and budgets, or worse, just duplicating the campaign and hoping for the best. This rarely works because it doesn’t address the root cause. Your underspending issue isn't the problem itself; it’s the most obvious symptom.

The real problem is that the delicate balance between your audience, your ad creative, and your campaign objective has broken down. You said "audience size the same, everything else same," but this is a common misconception. A custom audience is not a static list. It's a dynamic group of people. Every single day, new people enter the audience (e.g., visit your website) and others leave (e.g., their 180-day window expires). More importantly, the behaviour of the people within that audience changes. Someone who was interested last week might have bought from a competitor yesterday. Someone else might have seen your ad ten times and is now completely blind to it.

Meta's conversion algorithm is incredibly sophisticated. Its one and only job is to get you the most conversions for the lowest possible cost, based on your campaign objective. When it starts to dramatically pull back on spend, it’s sending you a very clear signal: "I've scanned the audience you gave me. Based on the historical performance of your ads and the recent behaviour of these users, I predict that I cannot achieve your desired conversion at an efficient price. Therefore, to avoid wasting your £1500 daily budget on clicks that won't convert, I am choosing not to spend it."

It's actually doing you a favour. I know it doesn’t feel like it, but it’s preventing you from burning cash. I remember one campaign we were running for a B2B software client saw a similar issue. Spend dropped off a cliff. We looked into the metrics and saw their ad Frequency was through the roof (over 8 in a 30-day period) and their CTR had tanked. The algorithm rightly concluded that showing the same ad to the same exhausted people was a waste of money. The solution wasn't to force the budget; it was to give the algorithm something new to work with.

So, our first job is to play detective and use Meta's own data to confirm this diagnosis. You need to look at these specific metrics in your Ads Manager for that ad set, comparing the last 7 days to the 7 days prior:

• Frequency: This is the average number of times each person has seen your ad. If this number is climbing, especially if it's above 3-4 for a warm audience, you're likely dealing with ad fatigue. People are tired of seeing your ad.
• First-Time Impression Ratio: This tells you what percentage of your daily impressions are going to people seeing your ad for the very first time. If this number is very low (e.g., under 20%), it means you are just hammering the same people over and over again. The well is dry.
• Click-Through Rate (CTR): Is your CTR declining? This is a classic sign of ad fatigue. An ad that was once compelling is now being ignored, leading to fewer clicks, which signals to the algorithm that the ad is no longer effective.
• Cost Per Mille (CPM): Are your CPMs (cost per 1,000 impressions) rising? As the algorithm struggles to find convertible users, it often has to bid on more expensive, less-engaged pockets of your audience, driving up costs before it eventually gives up and cuts spending altogether.

Looking at these numbers will almost certainly reveal the story. You haven't changed anything, but your audience has. They've become saturated. They've become fatigued. And the algorithm has noticed.

I'd say you need to change your perspective on 'awareness'...

A lot of people, when faced with this, think the answer is to switch their campaign objective to "Reach" or "Brand Awareness" just to force the budget to spend. This is a catastrophic mistake. When you select "Reach," you are explicitly telling Meta: "Find me the largest number of people for the lowest possible price, and I don't care if they ever click or buy anything."

The algorithm dutifully complies. It goes out and finds the users in your audience who are least likely to engage, least likely to click, and absolutely least likely to ever convert. Why? Because their attention is cheap. No other advertiser wants them for conversion campaigns, so their ad inventory is sold at a discount. You end up paying to show your ads to the worst possible segment of your audience.

A conversion campaign, like the one you're running, is infinitely smarter. It actively seeks out the people who exhibit behaviours associated with converting (e.g., they click on ads, they complete purchases, they fill out forms). When it stops spending, it's not because it's broken; it's because it's too smart to waste your money on people it knows won't act. True brand awareness for a growing business is a byproduct of effective direct-response advertising, not a goal in itself. The best way to build your brand is to get a customer, deliver an amazing experience, and have them tell their friends. That starts with a conversion.

Your job isn't to force the algorithm to spend. Your job is to give it new, better options that it can be confident in. This means a two-pronged attack: new creative and new audience structures.

You probably should refresh your creatives first...

This is your number one lever. Before you even touch your audience settings, you must address the ads themselves. The people in your warm audience have likely seen your current ads multiple times. They are now invisible. You need to shock them out of their banner blindness with something completely new.

Don't just change the headline or the button colour. You need to test fundamentally different angles. If you've been running an ad that focuses on a feature, run a new ad that focuses on the painful problem your product solves. We often use a framework called Problem-Agitate-Solve (PAS) for service businesses.

-> Problem: State the specific pain point your customer is feeling. "Are your warm audience ads suddenly refusing to spend their budget?"
-> Agitate: Pour salt on the wound. Describe the frustration and consequences. "You're checking Ads Manager ten times a day, watching your sales dry up, and have no idea what's broken or how to fix it."
-> Solve: Introduce your product or service as the clear, simple solution. "We diagnose the hidden saturation in your audiences and build a fresh campaign structure that forces the algorithm to find new pockets of customers. Get your ads spending and converting again."

For a product, especially SaaS, the Before-After-Bridge is powerful.

-> Before: Describe their current, frustrating reality. "Your ad budget is £1500/day, but only £25 is being spent. Your ROAS is collapsing."
-> After: Paint a picture of the ideal future. "Imagine confidently scaling your budget, knowing every pound is being spent efficiently to acquire new customers at a predictable CPA."
-> Bridge: Position your product as the bridge to get them there. "Our tool automatically identifies ad fatigue and suggests new creative angles, ensuring your campaigns never go stale."

Think about formats too. If you've only used static images, you MUST test video. It doesn't have to be a high-production affair. We've seen SaaS clients get incredible results with simple, authentic User-Generated Content (UGC) style videos—just a customer talking to their phone camera about how the product helped them. It cuts through the noise and builds trust instantly. Create 3-5 brand new ads that look and feel completely different from what you've been running. This is non-negotiable.

You'll need to restructure your audiences...

While new creatives are being developed, you can work on the second part of the puzzle: your audience targeting. Your current custom audience is tapped out. We need to find new ways to segment it or build new audiences that the algorithm hasn't seen before.

Here are some of the most effective strategies we use, which you should test in separate ad sets:

1. Shorten the Lookback Window: If you're using a 180-day website visitor audience, you're targeting a lot of people who visited your site six months ago and have long since forgotten about you. The algorithm has likely exhausted the tiny fraction of those people who were still interested. Create new audiences with much shorter windows: 30 days, 14 days, and even 7 days. These people are much 'warmer' and more likely to be in a buying mindset. The audience size will be smaller, but the quality will be far higher, which is exactly what the algorithm is looking for.
2. Create Engagement Audiences: Don't just rely on website visitors. Create custom audiences of people who have engaged with your Facebook or Instagram page in the last 30-60 days. These are people who have actively liked, commented, or saved your content. They've raised their hand to say they're interested. This is often a completely fresh pool of people for a conversion campaign to target.
3. Build Value-Based Lookalikes: If you have enough purchase data, this is the most powerful tool in your arsenal. Don't just create a lookalike of "all purchasers." Create a custom audience of your best customers (e.g., using a CSV upload of customers with the highest lifetime value). Then, create a 1% lookalike audience from that source. You are explicitly telling Meta, "Go find me more people who look exactly like my most profitable customers." This almost always outperforms a generic purchaser or website visitor lookalike. We worked on a campaign for a subscription box client and this exact tactic was what allowed them to scale, eventually hitting a 1000% Return On Ad Spend.
4. Test High-Intent Retargeting: Instead of targeting all website visitors, get more specific. Create an audience of people who "viewed content" or "added to cart" but didn't purchase in the last 14-30 days. This is your lowest-hanging fruit. They were *this close* to converting. A new ad with a compelling offer (like a small discount or free shipping) can be incredibly effective here and kickstart the algorithm's learning.

The goal here is not just to find more people, but to give Meta more distinct, high-quality signals. When you provide it with a fresh, highly-qualified audience and a brand-new creative, it has a much better chance of finding that magical combination that leads to efficient conversions, and it will start spending your budget with confidence again.

This is the main advice I have for you:

To put this all together, you need to move away from your current campaign and launch a new, structured testing campaign. Don't just edit your existing ad set; the algorithm already has a negative history with it. You need a clean slate.

I've detailed my main recommendations for you in a table below. This is a clear, actionable plan to not only get your ads spending again but to build a more resilient and scalable campaign structure for the future.

This kind of situation highlights why paid advertising is so much more than just setting up a campaign and letting it run. It's a constant process of monitoring, diagnosing, and iterating. The platforms change, your audience's behaviour changes, and your competitors are always evolving. Having the experience to correctly interpret the data and know which levers to pull (and in what order) is what separates campaigns that stall from those that scale successfully.

It can be a lot to manage, especially when you're also trying to run a business. If you'd like to have an expert pair of eyes look over your account and provide a more detailed, specific strategy, we offer a completely free, no-obligation initial consultation. We can walk through your campaigns together, analyse the data, and give you a clear roadmap for getting things back on track.

Regards,

Team @ Lukas Holschuh