Flat sales with rising spend is where most founders realise their Meta account is not underperforming by accident. This Facebook ad recovery case study is about that exact moment – when a Shopify brand had enough data, enough budget, and enough agency history to know the problem was not demand. It was execution.
The brand was doing real revenue, not playing startup theatre. They had a proven product, a healthy average order value, repeat purchase behaviour, and a Meta budget that should have been producing far better returns. Instead, results were volatile. One week looked decent, the next fell apart. Cost per purchase drifted up, spend allocation made little sense, and reporting was full of activity without much commercial clarity.
That pattern is common because most ad accounts do not fail in dramatic fashion. They decay. A few campaigns get layered on top of old ones. Audiences start overlapping. Creative gets refreshed without a real testing framework. Budget gets pushed into whatever looked best last week. Before long, the account becomes noisy, reactive, and hard to trust.
The account was busy, not effective
When we reviewed the account, the first issue was obvious. It had too much going on and too little logic holding it together. There were multiple campaign types running at once, inconsistent naming conventions, duplicated audience ideas, and budget spread so thin that meaningful learning was almost impossible.
This is where a lot of agencies hide. They call complexity sophistication. It usually is not. For a founder, complexity just makes poor performance harder to diagnose.
The second issue was creative strategy. There was no shortage of ads, but there was a shortage of angles. Different videos and statics were saying roughly the same thing in slightly different packaging. That gives the illusion of testing while producing very little new information. Creative fatigue set in quickly, and the account kept leaning on old winners long after they had stopped pulling their weight.
The third issue was optimisation discipline. The account had been managed around platform signals and short-term metrics rather than revenue quality. That meant too much attention on click-through rates, engagement, and cheap traffic, and not enough on contribution to actual sales growth. Plenty of agencies still do this because vanity metrics are easier to explain than a weak month.
What changed in this Facebook ad recovery case study
Recovery did not start with scaling tactics. It started with stripping the account back to a structure that could be trusted.
We rebuilt campaign architecture around clear roles. Prospecting needed room to find new buyers. Retargeting needed to stop cannibalising what was already likely to convert. Existing customer activity needed to be separated so it stopped flattering acquisition performance. That sounds basic, but basic is where most accounts break.
Once the structure was cleaned up, budget allocation became more rational. Instead of funding too many ideas at once, spend moved toward fewer campaigns with enough volume to generate usable learning. That lowered noise and made performance shifts easier to read. Founders do not need more dashboards. They need cleaner signals.
Creative was treated as the growth lever, not a garnish. We did not just swap formats. We rebuilt the testing around distinct hooks, buying objections, offer framing, and product education. One angle focused on the product outcome. Another handled scepticism directly. Another leaned into social proof without sounding desperate. The point was not to make prettier ads. The point was to find messages that could convert cold traffic at scale.
There was also a targeting reset. Not because targeting is dead, and not because it solves everything. It depends on the account, the market, and the product. In this case, audience structure had become cluttered enough to create internal competition. Simplifying it gave the algorithm more room to work while making spend distribution less erratic.
The real fix was operational, not cosmetic
A lot of so-called account turnarounds are really just reporting exercises. Spend gets reduced, branded traffic props up the numbers, and everyone pretends efficiency improved. That is not recovery. That is shrinking the problem.
In this case, the goal was stronger revenue performance without hiding behind lower ambition. That meant accepting a short period of controlled instability while the rebuilt account relearned. Founders need to hear this more often: a proper reset can temporarily make results look messier before they become more stable. Anyone promising instant recovery is usually selling rubbish.
The benefit of doing it properly showed up within the first few weeks. Volatility started easing because campaigns had clearer jobs. Creative testing produced more usable learnings because variables were cleaner. Budget decisions became less emotional because there was finally a structure worth backing.
By the end of the recovery window, the account was in a very different place. Revenue from Meta had lifted materially, cost efficiency had improved, and the team had better visibility into which messages were actually driving purchases. Just as important, the founder was no longer reacting to random swings with guesswork.
What the numbers usually hide
Most case studies cherry-pick a ROAS screenshot and call it proof. That is lazy.
A serious Facebook ad recovery case study should show what changed beneath the number. Did the account reduce customer acquisition cost because the offer got sharper, or because spend was pulled back? Did revenue grow from new customer purchases, or did retargeting just mop up demand that was already there? Did creative performance improve because the concept got better, or because frequency had not caught up yet?
Those distinctions matter if you are spending real money. A founder running a Shopify brand at scale does not need another agency talking about blended metrics as if they are magic. You need to know whether the Meta account is genuinely creating more demand, capturing more of it efficiently, or simply taking credit for sales that would have happened anyway.
That is why recovery work is less glamorous than scale content on social media. It requires discipline. You audit what is broken, cut what is unnecessary, rebuild what matters, and test with intent. No heroics. No nonsense. Just better decisions repeated consistently.
Why most recoveries fail
Most recoveries fail for two reasons. The first is impatience. Founders have already lost time and money, so there is pressure to force immediate wins. That often leads to frantic changes, overcorrection, and more instability.
The second is weak accountability. Agencies love saying they are data-driven until results turn ugly. Then the excuses start – seasonality, market conditions, attribution, creative fatigue, consumer sentiment. Some of those factors are real. None of them excuse sloppy account management.
A good operator can tell the difference between a hard market and a bad setup. If the market is tough, strategy adjusts. If the account is broken, it gets rebuilt. But someone has to own the outcome.
That is why founder-led ecommerce brands usually do better with a specialist partner who is comfortable being judged on revenue movement, not presentation quality. If an agency cannot explain exactly what is broken, what is changing, and how success will be measured, they are not recovering anything. They are buying time.
Underdog Marketing was built around that reality. Tight selection. Clear accountability. No results, no fee. Because if a Meta account is already spending thousands each month and still underdelivering, the founder does not need more theory. They need a fix that shows up in revenue.
What founders should take from this
If your Meta account feels unpredictable, that does not always mean the platform is the problem. Sometimes the issue is simpler and more frustrating. The account has become bloated, the creative has gone stale, and nobody has had the discipline to reset the system before spending more money into it.
Recovery is rarely about one clever tactic. It is usually about removing clutter, restoring signal quality, and making the account easy to trust again. Once that happens, scaling gets a lot less chaotic.
If you are already doing meaningful revenue on Shopify and spending enough on Meta to expect better, stop tolerating accounts that are merely active. Activity does not pay wages. A clean structure, stronger creative logic, and ruthless accountability do.
The useful question is not whether your ads can improve. It is whether the person managing them can tell you exactly why they are underperforming, and what they are prepared to own when they fix it.