Hi there,
Thanks for reaching out! Happy to give you some initial thoughts on the campaign troubles you're having. It's a really common situation to be in, especially with Meta's CBO, which can feel like a bit of a black box sometimes. It often seems like it has a mind of its own, and not always in a good way.
The short answer is that you're right to be concerned, and the platform is essentially taking your money and finding the cheapest, easiest people to show your ad to, not the people most likely to buy. We can definitly fix that, but it involves taking back a bit of control and giving the algorithm a much clearer set of instructions. Let's get into it.
TLDR;
- Your immediate problem is Meta's algorithm prioritising cheap reach over actual performance. You're paying it to find non-customers. You must turn off the underperforming ad immediately.
- Don't just wait and hope. A week of burning cash on a bad ad is a week wasted. You need to be proactive and kill what isn't working.
- The root cause isn't just CBO vs. ABO; it's your campaign structure. Lumping multiple creatives into one ad set is a recipe for the exact problem you're seeing. A structured testing approach is needed.
- You must ensure your campaign is optimising for Sales (conversions). If you're not telling Meta to find buyers, it won't. All other metrics like CTR and CPC are secondary to your Cost Per Purchase.
- This letter includes a flowchart explaining the algorithm's flawed logic and an interactive calculator to help you figure out your target Cost Per Acquisition (CPA).
We'll need to look at why Meta is doing this...
Alright, so the first question is why on earth Meta is pumping 75% of your budget into a single ad that's clearly performing terribly. It's frustrating, and it feels completely illogical from a business perspective. But from the algorithm's perspective, it's doing exactly what it's been told to do, especially in the early days of a campaign.
When you launch a new campaign with CBO, the system enters a "learning phase". Its primary job is to spend your budget as efficiently as possible while exploring where it can get results. The problem is, without a lot of data (like sales data from your pixel), its definition of "results" is very loose. Right now, it's not looking for sales. It's looking for the path of least resistance to spend your $45 a day. Your ad, Video5, with its high CPM and low CTR, has likely found a very cheap, very large, and very passive audience pocket, probably on Instagram Reels or Stories where people scroll endlessly without clicking. The algorithm sees it can get loads of impressions (reach) there for cheap, so it declares "Success!" and doubles down, starving your other, better ads of budget.
You're basically falling into a trap I see all the time. When you let the algorithm optimise for broad metrics like reach without strong conversion data, you are actively instructing it to find the worst possible audience for your product. You are paying the world's most powerful advertising machine to find you the worst possible audience for your product because those users are not in demand. Their attention is cheap.
It’s a common myth that you should just 'trust the algorithm' blindly from day one. You can't. You have to guide it, especially with a new pixel and a new product. You're the expert on your business, not the machine. Your job is to give it tight constraints and clear goals so its 'learning' heads in a profitable direction, not just a cheap one.
Your CBO Campaign Starts
Goal: Spend $45/day
Algorithm Tests Ads
Looks for easiest way to get impressions
Finds "Video5"
Gets cheap reach on Instagram Stories to a passive audience
Decision Point
"This is efficient!" (at getting impressions)
Action Taken
Allocate 75% of budget here. Starve other ads.
I'd say you need to take control, not wait...
This leads directly to your next questions about what to do. The answer is simple: you need to be decisive. You can't afford to let the algorithm burn your money for a week in the hopes it'll magically figure things out. It won't. It'll just keep doing what it's doing because, from its limited perspective, it's succeeding.
So, to be brutally honest:
- Should you leave the campaign for a week? Absolutely not. Turn off that "cancerous ad" (your words, but accurate!) right now. Today. Don't waste another dollar on it. The data is already clear. It has a high CPC, a terrible CTR, and it's spent the most money. It's a dud. Killing losing ads quickly is one of the most important skills in paid advertising.
- Should you turn off the ad? Yes. Immediately. This is non-negotiable.
- Should you do ABO? This is a more nuanced question. Switching to Ad Set Budget Optimization (ABO) is a perfectly valid tactic to regain control. It would allow you to force the budget where you want it to go—for example, giving $22.50 to your promising AdSet1 and $22.50 to AdSet2 (after you've turned off Video5 within it). This is a good short-term fix to ensure your better creatives actually get a fair shot.
However, I'd say that seeing this as just a CBO vs. ABO problem is a mistake. It's a sticking plaster on a bigger issue, which is your overall testing structure. Both CBO and ABO are just tools. A good builder can build a solid house with either a hammer or a nail gun, but a bad builder will make a mess with both. The real problem is the blueprint – your campaign structure.
| Budgeting Method | Pros (For Your Situation) | Cons (For Your Situation) |
|---|---|---|
| CBO (Campaign Budget Optimisation) | Good for scaling winning campaigns once you have proven ads and audiences. More hands-off long term. | Terrible for initial testing with unproven ads. The algorithm will bully the budget towards one ad, giving you false negatives on others. This is exactly your problem. |
| ABO (Ad Set Budget Optimisation) | Excellent for initial testing. Guarantees each ad set gets a specific budget, ensuring a fair test for your different creative approaches (videos vs. static images). Gives you full control. | Requires more manual management. You have to monitor performance and shift budgets yourself. Can be less efficient at scaling than a well-optimised CBO campaign. |
You probably should rethink your testing structure...
This is the heart of the matter. Your current setup is making it impossible to learn anything useful. You've put four videos in one ad set and six creatives (photos and videos) in another. When you do this, you're not actually testing them against each other in a controlled way. You're just throwing them all in a bucket and letting the algorithm pick a favourite based on its own flawed, early-stage logic. It will almost always pick one and ignore the rest, just as you've seen.
A much more effective way to test is to isolate your variables. The gold standard for creative testing, especially at the start, is to use a structure where you have one unique ad per ad set. It sounds like more work, but it's the only way to get clean data and know for sure what's working.
I would reccomend a structure like this, using ABO:
| Campaign Level (ABO, Sales Objective) | Ad Set Level (Budget: e.g. $10/day) | Ad Level |
|---|---|---|
| Creative Test Campaign 1 Audience: [Your Best Guess ToFu Audience] |
Ad Set 1: Test Video 1 | Ad: Video 1 |
| Ad Set 2: Test Video 2 | Ad: Video 2 | |
| Ad Set 3: Test Photo 1 | Ad: Photo 1 | |
| Ad Set 4: Test Photo 2 | Ad: Photo 2 |
With this setup, each creative gets a dedicated budget ($10/day in this example). After a few days, you'll have much clearer data. You'll be able to see which specific creative is getting you the best results (ideally, sales, or at least adds to cart) for the money spent. Then you can turn off the losers and put more budget behind the winners, or move the winning creative into a new CBO scaling campaign.
The other huge missing piece here is your audience. You haven't mentioned who you're targeting. Are these ad sets targeting the same people? Different people? You could have the best ad in the world (like your Video1 seems to be), but if you show it to the wrong audience, it'll fail. You need to be methodical about this. I usually prioritise audiences based on how close they are to making a purchase. For a new store, you'll be starting at the Top of Funnel (ToFu), which means targeting based on interests, behaviours, and demographics. For example, if you're selling handcrafted leather wallets, you'd target interests like 'leatherworking', 'Etsy', 'men's fashion', and brands that align with yours. You need to get inside your ideal customer's head. What do they like? Who do they follow? This is far more important than just letting Meta target broadly.
You'll need to focus on what actually matters: conversions...
You mentioned you've had no sales. This is the only metric that truly matters. Your focus on CPM, CPC, and CTR is good for diagnosing problems with an ad (e.g., a high CPM and low CTR on Video5 tells us it's not resonating), but they are not the goal. You can have a fantastic CTR and still go broke if no one buys.
This brings up the most important question: Is your campaign objective set to "Sales"? If it's set to "Traffic" or "Engagement" or "Awareness", you are telling the algorithm to find you people who click links or like posts, not people who buy things. This is a critical error many people make. You must optimise for the final action you want someone to take. If you want sales, you must select the 'Sales' objective. This tells Meta to use its data to find users within your target audience who have a history of making online purchases and are most likely to convert on your site.
Once you're optimising for sales, the key metric becomes your Cost Per Acquisition (CPA) or Cost Per Purchase. How much are you paying to get one sale? To understand what a 'good' CPA is, you need to know how much you can afford to spend. This is where understanding your numbers is vital. Your product is $80. What's your profit margin on that? Let's say your cost of goods is $30, so your gross margin is $50 per sale.
A healthy benchmark for ecommerce is often a 3x Return on Ad Spend (ROAS). This means for every $1 you spend on ads, you get $3 back in revenue. In your case, to get $80 in revenue, you could afford to spend around $26.67 on ads. That is your target CPA. Now you have a clear benchmark. If an ad is getting you sales for under $27, it's a winner. If it's spending $50 without a sale, it's a loser and you turn it off.
This is the math that separates professional advertisers from people just boosting posts. It lets you make cold, hard, data-driven decisions instead of emotional ones. I've built a small calculator for you to play with these numbers yourself.
What's Your Target CPA?
Your Maximum Affordable CPA is $26.67
This is the main advice I have for you:
To wrap this all up, you're not in a bad spot. You've correctly identified a problem early, and you have some creative that shows real promise. You just need a more professional and structured approach to give yourself a fighting chance. Here is a table summarising my main recommendations for you to implement.
| Phase | What To Do | Why It Matters |
|---|---|---|
| Immediate Actions (Today) | Turn off 'Video5' and any other ads with very poor metrics (high spend, no clicks/sales). | Stops you from burning cash on proven losers. Frees up budget for potential winners. |
| Short-Term Strategy (This Week) | Duplicate your existing campaign. In the new one, switch to ABO. Create one ad set for each of your best creatives. Give each a small, equal budget (e.g., $10/day). | This creates a controlled testing environment. It guarantees each ad gets a fair chance to prove itself so you can find your true best-performing creative. |
| Foundational Fixes (Crucial) | Ensure your campaign objective is set to 'Sales'. Research and implement specific, relevant interest targeting for your ideal customer. | This tells the algorithm to find buyers, not just clickers. Showing the right ad to the right person is the fundamental formula for success. |
| Ongoing Optimisation | Monitor performance based on Cost Per Purchase (CPA), not just CTR/CPC. Use the calculator above to know your target CPA. Turn off ads that exceed this without sales. | This shifts your focus to profitability. You will make data-driven decisions that directly impact your bottom line. |
Running paid ads, especially for a new store, is a minefield of costly mistakes. It's very easy to spend hundreds or even thousands of dollars and have nothing to show for it but confusing data. The difference between success and failure often comes down to having a methodical process, understanding the math behind profitability, and knowing how to interpret what the platform is telling you.
This is where getting some expert help can make a huge difference. Instead of spending months and a lot of your own money on trial and error, an experienced eye can spot these structural problems in minutes and get you on the right path quickly. I remember one campaign we worked on for a women's apparel store where we drove a 691% return; we've learned these lessons the hard way so our clients don't have to.
If you'd like to go over your setup in more detail, I'd be happy to offer you a free, no-obligation 20-minute strategy session where we can screen-share and I can give you some more specific pointers on your audiences and creative. Sometimes a quick chat is all it takes to get things moving in the right direction.
Hope this helps!
Regards,
Team @ Lukas Holschuh
