Solved: Daily Spending Limit Issues on Facebook Ads Account

Hi there,

Thanks for reaching out regarding the issues you're having with your new Facebook Ads account and wanted to give you some of my thoughts. It can be incredibly frustrating when the platform seems to have a mind of its own, especially when you're trying to get a new venture off the ground. The spending limit fluctuations, the weird CBO behaviour... tbh, it's all par for the course with new accounts, but that doesn't make it any less of a headache.

The thing is, these glitches are often symptoms of a much bigger challenge, which is trying to get results from a platform that feels like a black box. The real goal isn't just to figure out why your daily limit went from $50 to $139 and back again, but to build a campaign strategy that's so solid it can weather these storms and actually deliver predictable results. So, while I'll definitely cover the immediate problems you're seeing, I want to give you a bit of a wider perspective that's helped the clients we work with move past this initial chaos.

We'll need to look at the 'Weird' Algorithm Behaviour...

First off, let's tackle the things that are causing you the most grief right now. The daily spending limit going haywire is something I've seen countless times with new accounts. Meta puts these limits in place as a sort of fraud prevention measure and to build trust with a new advertiser. As you make consistent payments, they gradually increase it. The jump to $139 and then back down to $50 is definately odd, but it's likely just the algorythm recalibrating itself. It's an automated system and it's far from perfect. My advice here is to not worry about it too much. As long as you're paying your bills on time, it will stabilise and increase over a couple of weeks. It's annoying, but it's temporary.

Now, about the CBO (Campaign Budget Optimisation) and how it's spending your money. You mentioned it prioritised 'weird ad sets at the expense of cheaper and demonstrably more engaging ads'. This is probably the single biggest misconception people have about how Meta's system works. It’s a bit of a myth that the algorithm cares about engagement metrics like likes, comments, or even a low cost-per-click on its own. It only cares about one thing: the conversion objective you set for the campaign.

If your campaign is optimised for, say, 'Purchases', the algorithm will ruthlessly seek out the users it believes are most likely to buy, regardless of what it costs to reach them. That 'weird' ad set that looks more expensive to you might be reaching a small, but highly valuable, pocket of users who are about to pull out their credit cards. The 'cheaper', more engaging ad set might be getting loads of clicks and likes from people who love your creative but would never in a million years actually buy anything. The algorithm knows this. It has more data on user behaviour than any of us ever will. So when you see it spending more on an ad set with a higher CPA, it's often because it's chasing a higher quality of user that it predicts will convert.

This is all part of what's called the 'learning phase'. A new campaign, especially on a new account, is basically flying blind. It needs to spend money across your different ad sets to gather data and figure out who your customers are. This phase can look messy. It can look illogical. The key is to give it enough time and enough conversions (Meta says 50 conversions per ad set per week, but even getting close helps) to exit this phase. Every time you pause the campaign or make significant changes, you risk resetting the learning. I know it's hard when you see money being spent in ways that don't make sense, but you often have to trust the process for a solid week or two before judging performance.

And on your last point, about knowing if your ads are even running... you're right to be sceptical. The dashboard data, especially in the main overview, can be delayed. I've seen it take hours, sometimes even a full day, to update accurately. Don't trust the notifications or the summary panels entirely. The single source of truth is the 'Delivery' column at the ad set and ad level within Ads Manager. If it says 'Active', your ad is running or eligible to run. If it says 'In Review', 'Paused', or has an error, then that's the real status. I generally tell clients not to even bother looking at today's data. Wait until tomorow to analyse yesterday's performance for a more accurate picture.

I'd say you need to stop fighting the algorithm and start feeding it...

So, if the algorithm is this powerful, weird machine, how do you work with it instead of against it? You stop trying to outsmart it on a micro level (like tweaking daily budgets constantly) and start feeding it a clear, structured strategy. The quality of what you put in dictates the quality of what comes out.

Most people just throw a few audiences into a campaign and hope for the best. A far better approach is to structure your campaigns based on the marketing funnel. You've probably heard of ToFu, MoFu, and BoFu (Top, Middle, and Bottom of Funnel). This isn't just jargon; it's a logical way to seperate your audiences and message to them correctly.

Here’s how I would usually prioritise audiences for a Meta ads campaign, especially for an eCommerce or lead-gen account. This structure provides clarity for both you and the algorithm.

You should have seperate, long-term campaigns for each stage of this funnel. A prospecting campaign (ToFu) and a retargeting campaign (MoFu/BoFu combined if your budget is small). Inside your prospecting campaign, you can have different ad sets testing your different audiences—one for interests, one for a lookalike, and so on. This structure keeps things clean and tells the algorithm exactly who you're trying to reach with what message.

For a new account like yours, you'll start at the top. You don't have enough data for retargeting or powerful lookalikes yet. So you begin with detailed interest/behaviour targeting to feed the pixel data. You need to get at least 100 website visitors, and ideally more like 1,000, before your MoFu audiences are even viable. You need at least 100 purchases before a 'purchaser lookalike' will be any good. It's a process of building up your data assets. For one B2B software client, we generated 4,622 registrations purely through well-structured Meta ad campaigns, but it started with disciplined testing of interest-based audiences to build up that initial data pool.

You probably should rethink who you're actually targeting...

This brings me to the most important part of the entire process, and the bit that 99% of advertisers get wrong. Even with the perfect campaign structure, your ads will fail if you're targeting the wrong people. Your CBO is prioritising 'weird' ad sets because it's desperately trying to find a buyer within the audience you've given it. If that audience is garbage, the algorithm can't work miracles.

Forget the typical Ideal Customer Persona (ICP) that says "We target women aged 25-40 who like yoga and live in London". It's useless. It's a demographic, not a person with a problem. To get paid ads to work, you have to stop thinking about who your customer is and start thinking about what their nightmare is.

Your ICP isn't a person; it's a problem state. It's an urgent, expensive, maybe even career-threatening problem that your product or service solves. For a B2B SaaS client I worked with, their target wasn't just 'HR Managers'. Their target's nightmare was 'spending 20 hours a week manually chasing timesheets and risking a payroll error that gets the whole company paid late'. See the difference? One is a job title. The other is a story full of pain and frustration.

Once you define your customer by their nightmare, your targeting becomes so much clearer. Where does this person go to try and solve their problem?
-> What niche podcasts do they listen to?
-> What industry newsletters do they actually read?
-> What software tools do they already use and pay for?
-> What influencers or experts do they follow on LinkedIn or Twitter?
-> What specific Facebook groups are they members of?

This is your targeting list. Instead of targeting the broad interest "Human Resources", you target interests like "Society for Human Resource Management (SHRM)", followers of specific HR influencers, or people with interests in competitor software. These are much sharper signals for the algorithm. Doing this work upfront is non-negotiable. Without it, you're just burning money.

You'll need an offer they can't ignore...

Let's say you've got your structure right and your targeting is laser-focused on the nightmare. There's one final piece of the puzzle: your offer. This is the #1 reason campaigns fail. An amazing product pitched with a terrible offer will go nowhere. An offer is not just your product; it's the entire package you present in your ad—the creative, the copy, and the call-to-action.

Your ad copy needs to speak directly to the nightmare you identified. Two powerful frameworks for this are Problem-Agitate-Solve (PAS) and Before-After-Bridge (BAB).

Problem-Agitate-Solve: You state the problem, you pour salt on the wound by describing how bad it feels, and then you present your product as the solution. For example: "Are your cash flow projections just a shot in the dark? (Problem) Are you one bad month away from a payroll crisis while your competitors are confidently raising their next round? (Agitate) Get expert financial strategy for a fraction of a full-time hire. (Solve)"

Before-After-Bridge: You paint a picture of their current frustrating world (Before). You show them the aspirational world they want to be in (After). And you position your product as the vehicle to get them there (Bridge). For example: "Your AWS bill just arrived. It’s 30% higher than last month, and your engineers have no idea why. (Before) Imagine opening your cloud bill and smiling. You see where every dollar is going and waste is automatically eliminated. (After) Our platform is the bridge that gets you there. (Bridge)"

This kind of copy works because it connects on an emotional level. It's not about features; it's about solving a painful problem.

But the copy is useless if it leads to a high-friction call to action. The 'Request a Demo' or 'Contact Us' button is often a campaign killer. It asks your prospect to commit to a sales call with a stranger. It's a huge ask. Your offer's only job is to deliver a moment of value for free to earn the right to ask for their time or money later.

If you're a SaaS company, the gold standard is a free trial (no card details). Let them experience the "aha!" moment inside the product itself. If you're a service business, you must bottle your expertise into a valuable asset. A free audit, a checklist, a calculator, a short video course. For one of our clients selling courses, we generated $115k in revenue in under two months because the initial offer was a free, high-value webinar that solved a genuine problem for the audience first. You have to give value to get value.

This is the main advice I have for you:

I know this is a lot to take in, especially when you were just asking about a wonky spending limit. But the reality of paid advertising is that these small tactical problems are almost always rooted in a bigger strategic weakness. Fixing your strategy is how you fix the results for good. I've detailed my main recommendations for you below:

As you can see, turning a struggling ad account around involves a lot more than just fiddling with settings. It requires a deep, strategic approach covering audience psychology, copywriting, and technical platform knowledge. It's a full-time job, and trying to learn it all through trial and error can be an incredibly expensive process, not just in ad spend but in lost time and opportunity.

This is where getting some expert help can make a huge difference. We spend all day, every day inside ad accounts, running these kinds of strategies for businesses from early-stage software startups to established eCommerce brands. We've seen what works, what doesn't, and how to get past these frustrating initial hurdles much faster.

If you'd like, I'd be happy to offer you a free, no-obligation 20-minute strategy session. We could have a look at your account together, and I can give you some more specific, actionable advice based on what I see. It's the same principle as the offers I mentioned above—we'll provide some real value for free, and you can see if our approach feels like a good fit for you.

No pressure at all, but the offer is there if you think it would be helpful. Either way, I hope this detailed breakdown gives you a clearer path forward.

Regards,

Team @ Lukas Holschuh