How We Used ABO to Beat CBO for a Niche eCommerce Brand's Meta Ads

19 June

Meta’s CBO is the default for scaling. For many eCommerce brands, that’s the right call.

But sometimes, it’s the fastest way to burn your budget on the wrong audience.

We saw this with a client in a very specific niche. They had a great product, a passionate audience, and a Meta Ads account that was bleeding cash. The ‘best practice’ of Campaign Budget Optimisation was failing them, and they couldn’t figure out why. We took a different path.

The result was a 1.8x ROAS increase in 60 days, without changing the creative or the total budget.

Here’s how we did it.

The unique challenges of niche eCommerce Meta Ads

First, let’s define ‘niche’ in a way that matters for media buying. It isn’t just about having a small audience.

It’s a combination of factors. A highly specific product, often with a high price point. A long consideration cycle. An audience that can’t be found by typing a single interest into Meta’s targeting fields.

Think custom-fit cycling components, not cheap t-shirts. Or specialised tools for a specific craft, not mass-market homewares.

These factors break the standard Meta Ads playbook. Broad targeting becomes inefficient. The algorithm struggles to find enough conversion data to learn effectively. When I was scaling my own stores, I saw this firsthand. You know who your customer is, but telling the algorithm is the hard part.

Our client was a perfect example. They sell a high-ticket item to a dedicated hobbyist community. The total addressable market in Australia is maybe 200,000 people. When we started working with them, their previous agency had set up CBO campaigns targeting broad interests.

Meta’s algorithm did what it was told. It found the cheapest impressions and clicks within those broad audiences. But these were low-intent users. The budget was spent on people who were vaguely curious, not the serious buyers who understood the product’s value. Their CBO campaigns were stuck, and their cost per acquisition was unsustainable. This is a common problem we see, and our Meta Audit covers the same checks we run to identify these issues.

Why Campaign Budget Optimisation (CBO) fell short

Campaign Budget Optimisation is designed to be a portfolio manager. You give it a budget and a collection of ad sets, and it automatically allocates spend to the top performers. You can find Meta’s official explanation on their Business Help Centre.

In theory, this is efficient. The algorithm makes thousands of micro-adjustments a day, far more than a human could.

But for this niche brand, CBO was making the wrong decisions. We identified three specific failure points.

First, the budget was always sent to one or two ‘proven’ ad sets that were already saturated. Meta saw they had the lowest historical CPA, so it kept feeding them money. This led to rocketing CPMs and ad fatigue. The performance wasn’t just flat, it was declining.

Second, promising new ad sets were starved of budget. We’d launch a new creative or a new audience hypothesis. CBO would give it maybe $20 in spend, see a slightly higher initial CPA than the saturated ad sets, and shut it off. The test never had a chance to gather meaningful data.

Third, we had no granular control. We couldn’t force the system to spend a set amount on a specific test. If we believed in an audience, we couldn’t back it with a protected budget. We were fighting the algorithm instead of guiding it.

The account was stuck in a loop. It was optimising for a local maximum, ignoring potentially huge wins because the initial data didn’t look perfect. This is a trap I see many accounts fall into, and it’s a core reason we developed The 3-Tiered Meta Ads Account Structure for Consistent Scale to impose more order. For this client, the result was a ROAS hovering around 1.5x and a feeling of complete helplessness.

Our strategic shift to ABO testing Meta Ads

The definition of insanity is doing the same thing and expecting different results. CBO wasn’t working. So we turned it off.

This decision goes against a lot of conventional Meta Ads advice. But our diagnosis was that the algorithm lacked the nuanced understanding of the customer required in this niche. It needed a human operator to take the wheel.

We switched the entire account to Ad Set Budget Optimisation, or ABO.

ABO is simple. Instead of setting the budget at the campaign level, you set it for each individual ad set. This gives you direct control over how much is spent on each audience and creative test.

Our new structure was built for deliberate testing. We broke the campaigns down logically. One campaign for top-of-funnel prospecting, another for retargeting. Inside the prospecting campaign, we created distinct ad sets for each hypothesis.

Ad Set 1: Lookalike (1% Purchaser) - Budget: $100/day
Ad Set 2: Interest Stack (Competitor Brands) - Budget: $75/day
Ad Set 3: Interest Stack (Niche Publications) - Budget: $75/day
Ad Set 4: Creative Test A (Video) vs Creative B (Carousel) - Budget: $50/day

This structure meant we could guarantee spend for each test. If we wanted to spend $50 a day testing a new video, we could. The algorithm couldn’t override that decision and push the budget back to the saturated lookalike audience.

We were back in control. We could analyse performance ad set by ad set, making manual adjustments based on what the data was telling us. This is a fundamental part of our process. It’s not about abandoning data, it’s about combining it with operator experience to make smarter decisions.

Detailed performance: ABO beats CBO for ROAS

The change was immediate and significant. We kept the total daily budget the same to ensure a fair comparison. The only major variable we changed was the shift from CBO to ABO.

Here’s a snapshot of the numbers, comparing the 30 days before our change to the 30 days after.

Before: CBO (30 Days) - Total Spend: $18,000 - Total Revenue: $30,600 - Blended ROAS: 1.7x - Cost Per Acquisition (CPA): $124 - Conversion Rate: 0.75%

The account was barely breaking even on ad spend, and that’s before factoring in COGS. It was not a sustainable model.

After: ABO (First 30 Days) - Total Spend: $18,000 - Total Revenue: $57,600 - Blended ROAS: 3.2x - Cost Per Acquisition (CPA): $68 - Conversion Rate: 1.4%

We nearly doubled the return on ad spend. The CPA was cut almost in half. The business was suddenly profitable from its top-of-funnel advertising.

So what actually happened?

The ABO structure allowed us to properly fund the ‘Niche Publications’ interest stack. Under CBO, this ad set was constantly ignored because its click-through rate was slightly lower. But by forcing a $75/day budget, we discovered that while fewer people clicked, the ones who did were far more qualified. Their average order value was 25% higher and their conversion rate was double that of the ‘proven’ lookalike audience.

This was an insight the CBO algorithm completely missed. It was optimising for cheap clicks, not for high-value customers. By taking control, we were able to nurture an audience that looked less promising on the surface but delivered much better bottom-line results.

This is just one example, but we have a track record of finding these hidden opportunities. You can see more of our results across different verticals. The principle is always the same: use data, but don’t be a slave to the default settings.

Operator insights versus Meta’s automated budget optimisation

This case study challenges a common belief in the industry: that you should always trust the algorithm.

Meta’s automation is powerful. For brands with large audiences and high conversion volumes, CBO is often the best choice. The algorithm has plenty of data to learn from and can optimise effectively.

But for niche brands, new brands, or brands with limited data, human oversight is critical. An experienced media buyer brings context that the algorithm lacks.

They understand the customer journey. They know that a customer in this niche might see an ad today, research for two weeks, and then then finally buy. The algorithm just sees a long click-to-conversion window and might deprioritise that ad set.

Our Meta advertising team acts as the strategic layer on top of the automation. We think of CBO like a self-driving car on a clear, straight highway. It’s efficient and reliable. ABO is like taking manual control to navigate a winding mountain road in the fog. You need a skilled driver who can read the conditions and make nuanced decisions.

For this client, their market was that foggy mountain road. The pixel data alone wasn’t enough. Our operator’s understanding of the hobbyist community allowed us to identify the ‘Niche Publications’ audience as a high-potential target, even when the initial metrics didn’t look perfect to the algorithm. That’s the difference between just running ads and actively managing a media buying strategy.

ABO vs CBO lessons for your strategy

The lesson here is not that ABO is always better than CBO. It’s that you need to use the right tool for the job. Blindly following ‘best practices’ without understanding the context of your own brand is a recipe for wasted spend.

So, when should you consider testing an ABO approach for your Meta Ads?

Here are a few indicators we look for: * Your CBO campaigns are unstable, with performance swinging wildly from day to day. * You operate in a niche market with a small, specific audience and a long sales cycle. * You need to run controlled tests on new audiences or creatives, but CBO isn’t giving them a fair shot. * Your CBO campaigns consistently push the budget to one or two ad sets, leaving others with almost no spend.

If any of these sound familiar, it might be worth structuring a controlled test. Duplicate your existing CBO campaign and switch the duplicate to ABO. Set manual budgets for each ad set based on your strategic priorities and let it run. The data will tell you which approach works better for your account.

Ultimately, Meta’s tools are just that: tools. They don’t replace the need for a sound strategy and a deep understanding of your customer.