Hi there,
Thanks for reaching out!
I've taken a look at the campaign data you shared, and it’s a common situation to be in – you've got something that's working, but it's volatile, and the fear of scaling it and breaking it is very real. The way you are looking at the data day-by-day is probably causing more stress than it needs to.
The short answer is you absolutely should not kill this campaign. But you do need a better system for managing and scaling it. Let's get into what that looks like.
TLDR;
- Stop making decisions based on daily ROAS. The volatility is normal. You need to look at 7-day or 14-day rolling averages to see the real trend.
- Don't just double the budget to scale ('vertical scaling'). This often breaks campaigns. You need a structured approach using 'horizontal scaling'—testing new audiences.
- The most important piece of advice is to build a proper campaign structure with dedicated prospecting (ToFu) and retargeting (MoFu/BoFu) campaigns. This will stabilise performance and unlock scale.
- Your real North Star metric isn't daily ROAS; it's your Customer Lifetime Value (LTV). Knowing your LTV allows you to spend more confidently to acquire customers.
- This letter includes a rolling ROAS visualiser to help you see the bigger picture and an interactive calculator to help you figure out your LTV.
We'll need to look at why daily ROAS is the wrong metric...
First things first, let's talk about the numbers you've shared. You're looking at them every single day and making a judgement. That's a recipe for a headache and bad decisions. Ad platforms like Meta have inherent volatility. Performance will swing up and down based on auctions, audience pockets, and a dozen other factors you can't control. Your job isn't to react to every dip, but to manage the overall trend.
Your breakeven is 1.7-1.8. Let's look at your performance since you went to a $200/day budget:
- Aug 8: 2.83
- Aug 9: 1.93
- Aug 10: 7.26 (The monster day)
- Aug 11: 3.91
- Aug 12: 1.23 (The 'bad' day)
- Aug 13: 3.08
- Aug 14: 3.34
- Aug 15: 2.77
Yes, you had one day that was below breakeven. But look at the average. Over those 8 days, you spent $1600 and your average ROAS was a whopping 3.29. That's nearly double your breakeven. Killing a campaign that's performing at 3.29x would be a massive mistake. You're letting the fear from one bad day cloud the fact that the campaign is very profitable overall.
You need to smooth out this data to see the real picture. Instead of looking at daily performance, you should be looking at 3-day, 7-day, or even 14-day rolling averages. This shows you the actual trend and stops you from panicking when you have a single off day. I've put your data into a chart below so you can see what I mean. Notice how the 7-day average is much more stable and consistently profitable, even when the daily numbers jump around.
I'd say you need to rethink scaling...
You mentioned you were worried about doubling the budget on a $200 campaign. You're right to be cautious. Just cranking up the spend on a single ad set (what we call 'vertical scaling') is often the fastest way to ruin a good thing. When you do that, you force the algorithm to aggressively find new people, which usually means performance gets worse and costs go up as it explores less optimal pockets of your audience.
A much safer and more sustainable way to scale is 'horizontal scaling'. Instead of putting more money into the *same* ad set, you duplicate your winning ad set and target a *new*, similar audience. This lets you expand your reach without disrupting what's already working.
Here’s a simple framework for making scaling decisions:
- Is your campaign/ad set consistently profitable over a 3-5 day period?
- If yes, you have two choices:
- A) Scale Vertically (Cautiously): Increase the budget by a small amount, like 15-20%. Wait another 3 days, see if performance holds, and repeat. This is slow and steady.
- B) Scale Horizontally (Aggressively): Duplicate the ad set. Keep the winning ad and copy, but change the audience. For example, if you're targeting an interest in "Handcrafted Jewelry," you could duplicate it and target an interest in "Etsy Shoppers" or a Lookalike audience of your past purchasers. This is how you find new pools of customers.
- If the answer is no, don't scale. Go back and analyse what's wrong. Is it the creative? The audience? The offer?
I've mapped this out in a simple flowchart below. This should be your decision-making guide, not your daily ROAS number.
Creative fatigue? Audience saturated?
Increase budget by 15-20%
Duplicate ad set to new audience
You probably should structure your account for proper testing...
The root cause of your volatility and scaling problems is almost certainly your account structure. Right now, it sounds like you have one campaign doing everything. This is like trying to fish for every type of fish with a single net. You might catch something, but it's inefficient and unpredictable. Professionals structure their accounts to mirror the customer journey, typically using a Top-of-Funnel (ToFu), Middle-of-Funnel (MoFu), and Bottom-of-Funnel (BoFu) approach.
- ToFu (Top-of-Funnel / Prospecting): This is where you find new customers. The goal here is to reach broad audiences who have never heard of you before. You'd use interest-based targeting, broad targeting, and Lookalike audiences of your best customers. This is where you do your horizontal scaling.
- MoFu (Middle-of-Funnel / Engagement): This is for people who have shown some interest but haven't gone deep. You retarget people who have watched your videos, engaged with your social media posts, or visited your website but didn't view a product.
- BoFu (Bottom-of-Funnel / Conversion): This is your highest-intent audience. You retarget people who have viewed products, added to their cart, or started the checkout process but didn't buy. This is where you close the deal.
By splitting your efforts like this, you can allocate your budget more effectively and tailor your message to where the user is in their journey. Your BoFu campaign will almost always have the highest ROAS, but it's small. Your ToFu campaigns will have a lower ROAS, but they are what feeds your funnel and finds new customers to enable growth. You cant just rely on retargeting as the audience will be too small and you'll run out of people to target very soon. You need to always get new people in at the top of the funnel.
Here’s a sample structure you could implement. You'd have your main budget in the ToFu campaigns, constantly testing new audiences horizontally. A smaller budget would go to the MoFu/BoFu campaigns to convert the interest you've generated.
| Campaign (Objective: Sales) | Ad Set | Audience Detail | Suggested Budget Split |
|---|---|---|---|
| C1 - ToFu - Prospecting | Ad Set 1: Interests | Target specific interests related to your product (e.g., 'Shopify', 'eCommerce software', etc.) | 70-80% |
| Ad Set 2: LAL Purchasers | 1% Lookalike of all customers who have purchased in last 180 days. | ||
| Ad Set 3: Broad | No detailed targeting, let the algorithm find customers (only works with a mature pixel). | ||
| C2 - MoFu/BoFu - Retargeting | Ad Set 1: Website Visitors | All website visitors in last 30 days (exclude purchasers). Show testimonials, social proof. | 20-30% |
| Ad Set 2: Cart Abandoners | People who added to cart in last 14 days (exclude purchasers). Show an offer, reminder, or address common objections. |
One campaign we worked on for a women's apparel store, for instance, implemented this exact structure. They went from inconsistent results to a 691% Return on Ad Spend because they stopped gambling on one campaign and started systematically testing audiences in a proper funnel. This is the difference between hoping for results and engineering them.
You'll need a better way to measure success...
While a structured funnel is crucial, there's an even more fundamental shift you need to make. The real question isn't "what was my ROAS today?", but "what is a customer actually worth to my business over their lifetime?". This is your Customer Lifetime Value (LTV).
Once you know your LTV, your entire perspective on ad spend changes. That "bad day" with a 1.23 ROAS might have brought in two customers who will go on to spend hundreds or thousands of pounds with you over the next year. Suddenly, that day doesn't look so bad, does it? It looks like an investment.
Calculating a basic LTV isn't too complicated. You need three numbers:
- Average Revenue Per Account (ARPA): How much a typical customer spends with you per order/month.
- Gross Margin %: Your profit margin on that revenue.
- Monthly Churn Rate %: The percentage of customers you lose each month (for subscription) or the inverse of your repeat purchase rate (for eComm).
The formula is: LTV = (ARPA * Gross Margin %) / Monthly Churn Rate
A healthy business aims for at least a 3:1 LTV to Customer Acquisition Cost (CAC) ratio. So, if your LTV is £900, you can afford to spend up to £300 to acquire a customer and still have a very healthy business. This frees you from the tyranny of daily ROAS and allows you to make strategic, long-term investments in growth.
Use the calculator below to get a rough idea of your own LTV. Play with the numbers and see how small changes in retention (lower churn) or average order value can dramatically increase what you can afford to spend on ads.
Based on a 3:1 LTV:CAC ratio, your affordable Customer Acquisition Cost (CAC) is £400.
This is the main advice I have for you:
To wrap this up, you're not in a bad spot; you're just at a point where gut feel and daily checks aren't enough. You need a system. The good news is that you have a campaign that is clearly profitable and has potential. Now you need to build the right structure around it to protect that performance and scale it reliably.
Here is the actionable plan I would recommend you start implementing right away:
| Action | Why It Matters | Your First Step |
|---|---|---|
| Stop Daily Analysis | It leads to emotional, short-sighted decisions and unnecessary stress. You're missing the profitable long-term trend. | Commit to only checking your overall ROAS every 3 days. Use the 7-day rolling average as your main health metric. |
| Implement a Funnel Structure | Separates new customer acquisition from converting warm leads, leading to more stable performance and clear data on what's working. | Create two new campaigns: one for Prospecting (ToFu) and one for Retargeting (MoFu/BoFu). Move your current ad sets into this structure. |
| Adopt Horizontal Scaling | Allows you to grow your ad spend and find new pockets of customers without breaking your best-performing ad sets. | Identify your best ad set. Duplicate it and test one new Lookalike audience (e.g., 1% of website visitors). |
| Calculate Your LTV | This is your true north metric. It tells you how much you can actually afford to spend to acquire a customer, freeing you from low-ROAS fear. | Use the calculator above with your best estimates for ARPA, margin, and churn. Determine your affordable CAC. |
Implementing this framework takes a bit of work upfront, but it's the foundation for scaling from $200/day to $2,000/day and beyond without constantly worrying that you're about to lose all your progress. It turns advertising from a gamble into a predictable system for growth.
Of course, building and managing this kind of system can be complex and time-consuming, especially when you're also running a business. This is often where expert help can make a significant difference, not just in setting things up correctly but in continuously testing and optimising to push performance even further.
If you'd like to go over your account together on a call and map out a more detailed strategy, we offer a completely free, no-obligation consultation. We could dive into your specific audiences and creatives and give you a more tailored plan of action.
Hope that helps!
Regards,
Team @ Lukas Holschuh
