How a fragmented ad account was rebuilt from the ground up, and what happened to performance when it was.
There’s a moment every paid media professional dreads. You open a client’s Ads Manager for the first time, and the campaign list just keeps scrolling. That was the reality here. Over 660 campaigns. Two years of history. And not a single stable signal in sight.
The uncomfortable truth about your ad account
You’ve been running Meta Ads for two years. Someone sends monthly reports. The numbers look like something. But you’ve never quite been able to answer the one question that actually matters: is this actually working?
If that question makes you pause, keep reading. Because the problem almost certainly isn’t your creative, your budget, or your targeting. It’s your structure.
The symptom you recognize, the cause you don't
The visible signs are familiar: a new post goes up, a new campaign gets created. It runs for a few days. Then another post, another campaign. Over and over. It feels like an activity. It isn’t a strategy.
This is called the boost model, and it is one of the most expensive ways to waste a Meta Ads budget. Here’s the mechanism that makes it fail:
Every new campaign starts Meta’s algorithm from zero. The system enters a “learning phase” where it tests different audiences, placements, and times of day to figure out who will actually take action. This phase needs roughly 50 optimization events to stabilize. A campaign that lives for five days never exits from it.
”Creating a new campaign for every post is like hiring a new employee every week, and firing them before they learn the job.
The result: you are permanently paying for the least efficient part of the Meta auction. Your budget funds guesswork, not optimization.
What the account actually looked like
The account belonged to a shopping center brand. Two locations, years of paid social activity. When the restructuring began, this was the starting point:
Six hundred and sixty campaigns. Every one of them short-lived. Every one of them resetting the algorithm before it could learn anything useful. Budget dispersed across too many surfaces to measure, let alone optimize.
The account wasn’t being advertised. It was being boosted. Those are not the same thing.
Before: fragmented model
- 660+ campaigns over 2+ years
- New campaign per post, per boost
- Algorithm resets every few days
- Budget spread thin across dozens of parallel campaigns
- No stable optimization signal
After: structured model
- 61 campaigns over 10 months
- Campaigns organized by objective
- Algorithm runs long enough to learn
- Budget concentrated on performing campaigns
- Stable, compounding signal over time
What the restructured system delivered
Over the following ten months, 61 campaigns were built and run. That’s the entire output of the optimization period, not 61 per month, not 61 per quarter. Sixty-one total, each designed to live long enough to actually learn.
The results at the campaign level:
| Campaign type | Result | Cost per action |
|---|---|---|
| Instagram profile visits | 11,000–12,000 visits/campaign | ~€0.06/visit |
| Page likes (follower acquisition) | 3,000+ new followers/campaign | €0.21–0.23/follower |
| Engagement | 1,100–2,800 interactions/campaign | €0.07–0.11/interaction |
Total spend during the optimization period: €15,719 across 61 campaigns. Total reach: 2,492,959. Comparable scale, lower spend, and, critically, a system that was actively improving rather than perpetually resetting.
Those 3,000 followers per campaign? They chose to follow. They weren’t just scrolled past. That’s what a stable algorithm, given enough runway, can actually find for you.
Is your account doing the same thing?
The boost model persists because it’s easy and it looks like activity. But before you dismiss this as someone else’s problem, ask yourself three questions:
How many campaigns does your Ads Manager currently show?
If the answer is in the hundreds, and your account is less than three years old, you almost certainly have a structural problem.
01
When did you last look at how many of your campaigns actually exited the learning phase?
If you can’t answer this, your budget has been funding guesswork for months.
02
Can you point to a single campaign that has been running continuously for more than 30 days and improving?
If not, your account has no compounding learning, just perpetual resets.
03
If any of those questions produced discomfort rather than a confident answer, the good news is that this is fixable. The fix isn’t more budget, it’s structure.
Let’s look at your account together
If this sounds like your setup, one conversation is usually enough to identify where the waste is. No pitch, just a direct look at your structure and what it’s costing you.
