Meta ad testing mistakes costing eCommerce brands profit, expert warns

Many eCommerce brands using Meta for advertising are failing to scale profitably due to outdated approaches to campaign testing, according to Joshua Uebergang, Director at Digital Darts Marketing Agency. Uebergang has overseen substantial Shopify ad spend and observed first-hand how traditional testing methods can undermine Meta's optimisation systems.

Testing issues

Uebergang highlighted that a commonly used approach-rapidly launching numerous ad sets with dedicated budgets-forces Meta to allocate spend across unproven creatives. He said this drastically increases wasted spend and prevents the platform's algorithms from efficiently optimising.

"Brands using this method waste eighty to ninety percent of their test campaign budget on creatives that flop," said Joshua Uebergang, Director, Digital Darts.

Uebergang further observed that many brands prematurely call an ad a 'winner' after minimal investment, often just USD $50 to $100. He believes this is misleading because small-scale results do not translate at higher spend levels.

"A ten times ROAS at fifty dollars a day is not success. That type of result is a small snapshot that has not been tested at volume or under pressure," said Uebergang.

Metrics focus

According to Uebergang, brands need to move away from relying on the Return on Ad Spend (ROAS) metric during testing. Instead, he recommends focusing on four key indicators: CPM (cost per thousand impressions), total spend, CPA (cost per acquisition), and frequency. He stated that these figures paint a much clearer picture of ad performance at scale.

"These four tell you everything you need to know," said Uebergang.

Creative stability

Many marketers attribute declining ad performance to audience fatigue, but Uebergang challenges this assumption. He explained that what is often interpreted as fatigue is frequently the result of instability in ad sets and a lack of opportunity for the system to optimise.

"What many marketers call fatigue is your ads never getting a chance to stabilise and grow. I have had ads last for years because they were launched and tested correctly," said Uebergang.

Single campaign strategy

For brands aiming to scale, Uebergang recommends consolidating all testing and scaling within a single, broad campaign. This structure should include a winners ad set and several test ad sets. Each test ad set would contain one flexible creative with three to five variations. He asserts that flexible creative enables Meta's systems to automatically allocate spend towards the best-performing combinations, reducing manual oversight and budget waste.

He also notes the importance of process when moving a successful creative to the winners ad set. "If you edit the ad, Meta will consider it a new ad and reset its learnings," said Uebergang.

Budget adaptation

For brands with larger budgets, he suggests using the same post ID in an Advantage Plus Shopping campaign to leverage Meta's automation capabilities. Smaller spenders, with daily budgets between USD $100 and $1,000, should keep activity within a single campaign to maintain control and data consistency.

Problem-driven testing

Uebergang criticised the tendency to test many formats without clear objectives, describing the process as brands "throwing everything at the wall and hoping something sticks." He asserts that effective testing involves problem-based messaging targeted toward specific customer pain points.

He shared an example involving Hydramama, a hydration brand that shifted its approach from targeting athletes to addressing the needs of pregnant and postpartum women. By adjusting messaging and packaging, the brand was able to connect with new segments and address distinct problems.

Efficiency and reach

Ultimately, Uebergang recommends every new campaign or variant aim to either enhance efficiency or expand reach. He warns that tests not delivering on these objectives are simply generating 'noise'.

"Good ads get fed, bad ads get starved," said Uebergang.