A ROAS recovery campaign example is more useful than another screenshot of a lucky week. If your Shopify brand is spending $3,000 or more a month on Meta and results have slipped, you do not need a fresh set of interest audiences and a cheerful report. You need to find where revenue is leaking, stop funding the leak, and give Meta enough clean signal to find buyers again.
This is what a real recovery can look like for an established ecommerce brand. The numbers are illustrative, but the operating logic is the same: diagnose first, rebuild the parts that are broken, and only scale once the account is producing profitable, repeatable sales.
The starting point: ROAS had fallen, but spend had not
Consider a founder-led Australian Shopify skincare brand doing roughly $1.8 million annually. Meta spend sat around $45,000 per month. Six months earlier, the account had produced a 3.2x blended Meta ROAS. It was not spectacular, but it supported growth.
Then performance deteriorated. Over the previous 30 days, Meta ROAS fell to 1.74x. Cost per purchase climbed from $38 to $61. The agency response was familiar: launch more campaigns, duplicate ad sets, turn targeting knobs, and call the result “testing”.
The founder had 17 live campaigns, 46 ad sets, and more than 80 ads. Most had spent too little to tell a useful story. Budget was fragmented. The account was optimising around inconsistent conversion data. Creative was tired, but nobody had a system for replacing it. Worse, reporting celebrated click-through rate while contribution margin was being chewed up.
That is not an audience problem. It is an account management problem.
A lower ROAS does not automatically mean Meta has stopped working. It may mean your offer has weakened against the market, your creative has been seen too often, your product page is leaking conversion, or your campaign structure is starving winning ads of spend. The job is to separate those issues before changing everything at once.
The ROAS recovery campaign example: what changed first
The first seven days were not about scaling. They were about removing noise.
We would start with an account audit across the previous 90 days, then compare it with the prior 90-day period. That means looking beyond platform ROAS: new customer revenue, average order value, conversion rate, repeat purchase behaviour, refund rate, gross margin and the gap between Meta-reported sales and Shopify sales.
For this example, the key findings were blunt. The brand’s best-performing ads were nearly four months old and frequency had pushed past a healthy range. Prospecting spend was split between broad, multiple interest stacks and a lookalike campaign with no meaningful incremental result. Retargeting had been given too much credit because it was mostly harvesting customers who were already close to buying.
The account also had a hidden commercial problem. A first-order discount had trained customers to wait for an offer, reducing margin on sales that may have happened anyway. A 2.8x ROAS on a discounted bundle is not necessarily better than a 2.2x ROAS on a full-price hero product. Founders who only chase the prettier dashboard number often scale the wrong thing.
Campaign architecture was simplified
Seventeen campaigns became three practical functions: prospecting, retargeting and controlled testing. The goal was not minimalism for its own sake. It was to consolidate conversion volume so Meta could learn from enough purchase events, while making budget allocation obvious.
The prospecting campaign carried the majority of spend and used broad targeting with sensible location, age and exclusion settings. Retargeting was capped to prevent it consuming disproportionate budget. The test campaign gave new concepts a defined place to earn spend without destabilising the main acquisition engine.
That structure will not fit every account. A brand with separate product categories, distinct customer groups or a high enough spend level may need more separation. But when a $45,000 monthly account has been chopped into dozens of tiny bets, consolidation is usually the fastest way to regain control.
Creative was treated as the recovery lever
Most underperforming Meta accounts do not have a targeting shortage. They have a creative shortage.
The skincare brand had polished studio images, founder content and product demos. The problem was repetition. Every ad made the same claim in slightly different clothing: clean ingredients, glowing skin, premium formulas. None addressed the immediate reasons a frustrated buyer would change their routine.
The recovery plan created new angles around specific customer tensions: makeup pilling over dry skin, morning routines that take too long, sensitive skin reacting to active ingredients, and the cost of buying products that sit unused in the bathroom cabinet. Each angle had multiple opening hooks, formats and proof points.
The team did not test random variations. They tested a clear hypothesis. For example, if the existing buyer was anxious about irritation, a creator-style video showing texture, application and a credible sensitive-skin routine could outperform another glossy brand film. If a bundle was driving higher average order value, creative needed to explain why the products belonged together rather than simply showing a discount badge.
New creative entered the test campaign first. Ads that generated efficient purchases and held conversion quality earned their way into prospecting. Ads that produced cheap clicks but weak sales were cut. That sounds obvious. Plenty of agencies still keep them because click metrics make the weekly report look less embarrassing.
Budget decisions followed profit, not panic
In week one, spend was reduced from $45,000 to $36,000 per month equivalent. This was not a retreat. It stopped the account from feeding budget into fatigued ads while new creative gathered evidence.
By the end of week two, the rebuilt prospecting campaign was producing purchases at $47, down from $61. Meta ROAS had lifted to 2.25x. More importantly, Shopify showed that the full-price hero-product offer had a stronger contribution profile than the discounted bundle.
At that point, budget increased in controlled increments. Not a 50 per cent jump because one ad had a good Tuesday. Spend was moved towards creative concepts that had held up across several days, enough purchases and more than one audience condition.
Over the next six weeks, monthly spend returned to $45,000 and then reached $52,000. Meta ROAS settled around 2.85x. New customer revenue increased by 34 per cent versus the recovery period baseline. The brand was not chasing a magical 5x platform number. It was generating more profitable revenue with a cleaner account and a stronger pipeline of creative.
What this example leaves out on purpose
A campaign rebuild cannot rescue a broken commercial model. If your landing page converts at 0.7 per cent, your stock is unreliable, your product has weak market fit, or your gross margin leaves no room for acquisition, Meta optimisation will not fix it.
Likewise, do not confuse attribution with reality. Meta can overstate retargeting results, particularly when a brand has strong email, organic and branded search demand. Use platform data to make daily media decisions, but check Shopify revenue, customer cohorts and margins to judge whether the spend is genuinely creating growth.
There is also a timing trade-off. Cutting poor spend fast protects cash, but changing every campaign, offer, audience and landing page on the same day destroys your ability to identify what worked. Recovery needs urgency, not chaos.
When a recovery campaign is the right move
A recovery campaign makes sense when your brand has a proven offer, enough conversion volume and a history showing Meta can produce customers profitably. It is particularly relevant when results declined after creative fatigue, account bloat, unmanaged scaling or an agency that spent more time explaining volatility than fixing it.
It is not the right answer if you are spending too little for reliable data or you expect Meta to compensate for a product customers do not want. No results, no fee sounds appealing until the numbers reveal there is no viable engine to improve.
For qualified Shopify brands, the opportunity is usually not hidden in another dashboard metric. It sits in the discipline to cut waste, make better ads, simplify decisions and hold the people managing your money accountable. That is the standard Underdog Marketing works to, because a founder cannot pay suppliers, staff or themselves with vanity metrics.
Before you approve another month of “testing”, ask one direct question: which specific change is expected to create more profitable new-customer revenue, and what evidence will prove it? If nobody can answer that clearly, the recovery should start there.