Solved: ROAS Goal vs Bid Caps on High Budgets

Hi there,

Thanks for reaching out! Happy to give you some initial thoughts on your question about bid caps versus the ROAS goal bidding strategy. It's a really good question, and one that gets to the heart of how to manage Meta's algorithm, especially when things feel a bit unstable.

You're spot on that they work differently, and your reasoning is pretty much correct. You absolutely shouldn't treat a ROAS goal campaign the same way you'd treat a bid cap campaign by setting a massive budget. I'll get into the weeds on why that is and what you should do instead. The short answer is you're better off starting with a realistic budget and a slightly conservative ROAS goal, and then scaling methodically from there. It's less about trying to trick the algorithm and more about giving it clear, achievable instructions.

But the bigger picture here is about creating stability in your account, and that goes a bit deeper than just the bidding strategy you choose. It's about your campaign structure, your understanding of your business's numbers, and your overall strategy. I'll walk you through all of it.

You're right to be cautious... Bid Caps and ROAS Goals are completely different animals

Okay, so let's get this first bit cleared up, because it's the most important part of your question. Your instinct is correct: the logic for setting a huge budget on a bid cap campaign doesn't apply to a ROAS goal campaign. In fact, it can have the opposite effect.

Think of it like giving instructions to an employee.

With a Bid Cap, you're saying: "Go out and find me as many customers as you can, but whatever you do, do not pay more than £20 for any single auction." If you give them a small budget, say £100, they might only find four or five opportunities that day that fit under the £20 cap. But if you give them a massive budget, say £10,000, you're not telling them to spend it all. You're just giving them the freedom to explore every possible auction during the day, confident that they'll only act on the ones that meet your strict price ceiling. The high budget unlocks more opportunities for the algorithm to assess against your rigid rule. It won't spend the £10k; it'll just use the flexibility to find all the sub-£20 auctions it can.

Now, with a ROAS Goal (or more accurately, Minimum ROAS), the instruction is very different. You're saying: "Go out and find me customers, and I want you to make sure that for every £1 I give you, you bring me back at least £4 in revenue. Here is your budget for the day."

This is where your reasoning is bang on. If you set a low budget, say £100, and a 4x ROAS goal, the algorithm thinks, "Okay, I need to find £400 of revenue from this £100. That's achieveable. I'll focus on the absolute lowest-hanging fruit - probably some warm retargeting audiences - to hit that goal."

But if you give it a massive budget, say £10,000, and the same 4x ROAS goal, the algorithm's calculation changes dramatically. It thinks, "Okay, they want me to spend £10,000 and bring back £40,000 in revenue... today. Based on the available audiences and likely conversion rates, that's almost impossible. The pool of people who will deliver a 4x return isn't that big. To spend the full £10k, I'd have to start bidding on much less likely converters, which would drag the average ROAS way down. Since my primary instruction is to not fall below 4x, I'd better just not spend the money."

And so, it throttles delivery. Your campaign underspends massively, not because it's saving you money, but because you've given it a set of contradictory and seemingly impossible instructions. You've told it to sprint a marathon. This is one of the most common reasons I see ROAS goal campaigns fail to spend their budget.

So how should you actually use a ROAS goal campaign?

The goal isn't to shock the system with a massive budget. The goal is to feed it good data and scale methodically. Here’s the process I'd typically follow, which focuses on stability and profitable growth rather than volatile experiments.

1. Set a Realistic Starting Point: Forget 10x. Start your new campaign with a daily budget that's in line with what you'd normally spend, or what one of your better ad sets is spending. Maybe 1x to 1.5x your average daily spend.
2. Set an Achievable ROAS Goal: This is where people often shoot themselves in the foot. Don't set your goal to the ROAS of your best-ever day. Look at your account's average ROAS over the last 30 days. Let's say it's 3.5x. I would recomend setting your initial ROAS goal at around 3.0x or 3.2x. Give it a target that is slightly easier to achieve than your current average. This encourages the algorithm to actually spend the budget, get conversions, and exit the learning phase faster. You can always increase the goal later once it's stable.
3. Be Patient: A new campaign needs time. Let it run for at least 3-4 days, ideally a week, without touching it. It needs to get through the learning phase. If it's underspending, your ROAS goal is likely too high for the audience/creative combination. If it's spending and hitting the goal, great.
4. Scale Methodically: Once the campaign is stable and consistently hitting (or exceeding) your ROAS goal, you can start to increase the budget. The golden rule is to not make drastic changes. Increase the budget by no more than 20-25% at a time, and then wait another 2-3 days to see how the system adapts. A small, steady increase is far less likely to shock the algorithm and reset the learning phase than a sudden doubling of the budget.

This process is boring. It's not a hack. But it's how you build stable, predictable performance, which sounds like exactly what you're after to ride out the "instability."

I'd say you need to prepare for the scaling plateau

Now, let's talk about that "instability". What you're likely experiencing is something that hits every single successful ad account. It’s the scaling plateau. And understanding this is far more important than any bidding tactic.

The logic is simple. When you have a low budget, Meta's algorithm finds the easiest, cheapest, most likely people to convert. These are your superfans, your warm retargeting audiences, people who have been waiting for a reason to buy. They deliver a fantastic ROAS.

But as you tell Meta to spend more money, it has to go looking for more people. It's already converted the easy ones. So, it has to move outwards to slightly less interested audiences. Then to lukewarm audiences. Then to colder audiences who might never have heard of you. With each step, the cost to acquire a customer goes up, and therefore your ROAS goes down. This is not a sign of failure. It is the natural law of paid advertising. You cannot scale indefinitely and maintain the same efficiency. It's impossible.

I remember one software client we worked with who was getting trials for about $7 each on a small budget with Meta ads, which was an incredible result. Naturally, they wanted to scale up aggressively. We had to manage expectations that as we increased the spend, that cost per trial would inevitably rise. And it did. As we scaled the campaign and started reaching new, colder audiences, the cost per trial increased, but were they less profitable? No. They were acquiring hundreds more customers every month, and their overall revenue growth was massive because they understood the trade-off between efficiency and volume.

The instability you feel is likely you pushing up against the edges of your current audience pool at your desired ROAS. The platform is trying to spend more, but it can't find enough people who will convert at a high enough value to maintain your historic ROAS, so performance fluctuates. The solution isn't a bidding hack; it's a strategic shift.

You'll need to know your numbers, especially your LTV

This leads to the most important question in all of advertising, and it’s not "What should my ROAS be?". It's "How much can I afford to pay to acquire a customer?". The answer to this is your Customer Lifetime Value (LTV).

If you don't know this number, you are flying blind. You're optimising for a vanity metric (ROAS) without knowing what return you actually need to be profitable. Let's break it down with a simple example.

• Average Revenue Per Account (ARPA): How much a customer pays you per month. Let's say £200.
• Gross Margin %: Your profit margin on that revenue. Let's say it's 70%.
• Monthly Churn Rate: The percentage of customers you lose each month. Let's say it's 5%.

The calculation is: LTV = (ARPA * Gross Margin %) / Monthly Churn Rate

So, LTV = (£200 * 0.70) / 0.05 = £140 / 0.05 = £2,800.

In this scenario, each customer you acquire is worth £2,800 in gross margin to your business over their lifetime. A healthy, sustainable business model often aims for a 3:1 LTV to Customer Acquisition Cost (CAC) ratio. This means you can afford to spend up to £2,800 / 3 = £933 to acquire a single customer.

Now think about your ROAS goal. If a new customer makes an initial purchase of £200, to get a CAC of £933 you would need a ROAS of just... £200 / £933 = 0.21x. This seems crazy low, but this is for a subscription service. If your business isn't a subscription service, LTV might be harder to calculate, but it's still about repeat purchase rate and average order value over time.

My point is, once you know you can afford to spend £933 on a customer, you stop panicking when your ROAS on the initial purchase dips from 5x to 4x. You realise you have a massive amount of room to play with. You can afford to bid more aggressively, reach broader audiences, and scale your business in a way that someone obsessed with maintaining a 5x ROAS simply can't. They'll stay small forever, while you grow intelligently.

I'd say you need to structure your account for stability

Finally, if you want to ride out instability, you need a structure that's built for it. Relying on one or two campaigns is risky. When they fluctuate, your whole account fluctuates. A better approach is to structure your campaigns around the marketing funnel: Top of Funnel (ToFu), Middle of Funnel (MoFu), and Bottom of Funnel (BoFu).

• ToFu (Top of Funnel - Cold Audiences): This is your prospecting. You're reaching people who have never heard of you. Here, you'll use lookalike audiences and detailed interest/behavioural targeting. The goal isn't immediate high ROAS; it's to feed the funnel. Your ROAS here will naturally be lower.
• MoFu (Middle of Funnel - Warm Audiences): These people have shown some interest. They've watched your videos, engaged with your page, or visited your website but didn't take a key action. You retarget them with different messaging, perhaps showing testimonials or case studies. Your ROAS here should be better than ToFu.
• BoFu (Bottom of Funnel - Hot Audiences): These are your hottest prospects. They've added to cart, initiated checkout, or viewed specific high-value pages. You retarget them with direct offers, reminders about their cart, or scarcity messaging. This campaign should have your highest ROAS. I often see people combine MoFu and BoFu if their budget is smaller, which is perfectly fine.

By splitting your campaigns this way, you create a more stable system. Your BoFu campaign will consistently generate high returns from your hottest prospects. Your ToFu campaigns will constantly find new people to turn into warm and hot prospects. If one ToFu audience starts to fatigue, you can swap it out without wrecking your entire account's performance, because your MoFu and BoFu campaigns are still there, converting the people your other prospecting efforts have already found. This gives you isolated variables to test and creates a much more predictable and scalable machine.

I've detailed my main recommendations for you below:

So to wrap up, the issue isn't just about bid caps vs. ROAS goals. The instability you're feeling is a classic symptom of scaling. By understanding the mechanics of how these bidding strategies actually work, shifting your focus to your core business metrics like LTV, and building a more resilient account structure, you'll be much better equipped to manage it.

It can be a lot to take in and implement correctly. Getting this kind of strategic framework right is often where expert help can make a huge difference, moving you from constantly reacting to platform changes to proactively building a system that can weather them. We specialise in exactly this sort of thing.

If you’d like to have a chat and walk through your account together, we offer a completely free, no-obligation 20-minute strategy session where we can look at your specific setup and give you some more tailored advice. It might be a good next step to get some clarity.

Regards,

Team @ Lukas Holschuh