The Italian Testing Machine: How Nine Brands Deployed 10,000 Micro-Budget Ads on Instagram
An analysis of 10,000 Instagram ads in the Italian market reveals a brutal high-velocity testing strategy. Nine brands deployed 538k EUR by force-feeding the algorithm micro-budgets and killing failing permutations at just 23 EUR.

The preciousness of the hero creative is dead in performance marketing. Modern media buying on Instagram looks less like a targeted sniper shot and more like a massive scattergun blast. Operators are no longer relying on a handful of polished video assets to carry their quarterly targets. Instead, they are pushing thousands of minor variations into the feed, letting the algorithm sort the winners from the losers in real time.
We recently pulled the data on a highly specific cohort of Instagram ads targeting the Italian market over a 90-day window. The numbers reveal a distinct playbook utilized by a very small group of aggressive operators.
The Scale of the Operation
Looking at a cohort of 10,000 Instagram ads in the Italian market, we can see the sheer volume of creative throughput. This is not a broad market average. This specific 10,000-ad cohort is driven by just nine brands.
Together, these nine brands deployed a total spend of 538,322 EUR over a 90-day period. On average, that is roughly 60,000 EUR per brand. A solid quarterly budget, but not necessarily a whale-tier enterprise spend.
What makes this cohort fascinating is the division of that spend. To push 538,322 EUR, these nine brands did not use a standard single-ad-account structure. They spread their operations across 31 distinct ad accounts.
Why does a single brand need over three ad accounts to run Instagram campaigns? Operators at this edge of the market use multi-account architectures for three primary reasons:
- Sandboxing: One account is dedicated purely to testing. It is the burn account where thousands of variations are uploaded and tested with micro-budgets. The winning creatives are then duplicated into a primary scaling account. This prevents the constant influx of new creatives from resetting the learning phase on the primary pixel.
- Objective Isolation: Different accounts are trained for entirely different outcomes. One account might be optimized purely for top-of-funnel reach, while another is trained exclusively on lower-funnel purchase events.
- Risk Mitigation: When you are pushing over 1,000 ads per brand in a 90-day window, the automated compliance bots will inevitably flag false positives. Operating across multiple accounts ensures that if one account catches a temporary suspension, the brand does not lose its entirely daily revenue stream.
The Brutal Efficiency of the Kill Switch
The most striking metric in this cohort is the median spend per ad. Despite a total spend exceeding half a million euros, the median spend sits at a minuscule 23.78 EUR. The 75th percentile of spend only reaches 30.54 EUR.
These operators are not giving campaigns three weeks to exit the learning phase. They are launching massive batches of dynamic creatives, dynamic text, and varied hooks. They spend exactly enough to get a preliminary signal, usually around 20 to 25 EUR, and immediately cut funding to anything that fails to gain traction.
This high-velocity testing breaks the standard reporting tools. Out of the 10,000 ads in this Italian cohort, only 161 ads managed to register a verifiable CPM. The platform categorizes the format of these high-churn ads as unknown, simply because they do not live long enough, or spend enough, to populate the standard transparency data points.
The Economics of the Survivors
When an ad does survive the 23 EUR kill switch, the unit economics are highly favorable. For the fraction of ads that lived long enough to register standard delivery metrics, the median CPM was recorded at 1.77 EUR.
The bottom quartile achieved an astonishingly cheap 0.46 EUR CPM, while the top quartile paid 10.04 EUR. In a Western European market like Italy, acquiring a thousand impressions for under two euros is a clear signal that the underlying creative resonated deeply with the algorithm and the audience. The platform rewards high engagement with cheaper distribution. By churning through thousands of permutations, these brands are brute-forcing their way to the cheapest possible inventory.
The reach metrics tell a similar story of efficiency. The median total EU reach for these ads was 4,184 users. Securing over four thousand unique eyeballs for a median spend of 23.78 EUR translates to highly effective brand penetration, even if the primary goal is direct response.
Building the Pipeline
Executing this playbook requires more than just aggressive media buying tactics. It requires a fundamental restructuring of the creative production pipeline.
If a single brand is pushing roughly 1,100 ads per quarter, they are launching roughly 12 new ads every single day, seven days a week.
This cannot be done with traditional agency retainers or standard video editing workflows. Brands executing at this level are leveraging heavily modular asset creation. They shoot a base video, but they test ten different text overlays within the first three seconds. They test five different audio tracks. They test the same video with a static image thumbnail versus a fast-motion thumbnail.
Every single variable is isolated, uploaded, and fed to the machine. The ones that fail cost 23 EUR. The ones that succeed buy cheap reach at a 1.77 EUR CPM and fuel the scaling accounts that drive the actual revenue.
For marketing operators looking to compete in saturated markets, the lesson is clear. The days of betting your monthly budget on a single, highly produced video are behind us. The winners are building testing machines, separating their ad accounts, and letting the brutal math of the 23 EUR kill switch dictate their creative strategy.
Keep Reading
Frequently Asked Questions
Start with one monitor. Free.
Add a brand, paste a couple of competitor handles, and see your first calibrated readout in under five minutes.