The Barbell Budget: Why Retail Marketplaces Cap Video Tests at 24 Euros but Pour Thousands into Static Images

Lead generation in the retail marketplace sector is a notoriously difficult game. When your objective is acquiring high-quality sellers or high-lifetime-value buyers, the top of the funnel can quickly fill with low-intent clicks. Operators combat this by testing relentlessly. We recently analyzed a distinct cohort that reveals exactly how media buyers are structuring these tests, and the data points to a highly polarized "barbell" approach to budget allocation.

We examined a cohort of retail marketplace lead generation campaigns run over a 120-day window. The dataset includes 165 ads deployed by 4 brands across 63 distinct ad accounts. The headline metric is the spend distribution, which shows a fascinating anomaly. The 25th, 50th, and 75th percentiles for spend all sit at exactly 23.78 Euros.

This flatline is not a coincidence. When three quartiles of your ads show the exact same spend, it indicates a strict automated rule or a rigid media buying playbook. These operators are spinning up dozens of creatives and aggressively cutting them the moment they hit a low threshold if they do not show immediate traction. But while the vast majority of ads live in this micro-test sandbox, a select few are given massive budgets.

The Format Divergence: Video vs Static

The real story emerges when we split these ads by format. We see a massive divide in how operators treat video compared to static imagery.

Video ads in this cohort see a median spend of 24.54 Euros. They are stuck firmly in the testing sandbox. Meanwhile, the static image ads tell a completely different story. The median spend for an image ad is a staggering 2,354.33 Euros.

Why are operators killing video ads at 24 Euros while letting static images run into the thousands? Look at the cost per thousand impressions (CPM). Video ads in this cohort boast an incredibly cheap median CPM of 0.44 Euros. In an era where CPMs regularly clear the 5-Euro mark, a 44-cent CPM feels like a lottery win.

But cheap reach does not equal intent.

The Intent Gap and The 2,205 Reach Plateau

Video is inherently a passive medium. Users scroll, a video auto-plays, and the platform registers an impression. Sometimes users click through out of curiosity, generating a cheap lead that ultimately fails to convert into a transacting user on the marketplace.

Static images force the user to stop, read the copy, and make a conscious decision to engage. The friction is higher, which acts as a natural filter for intent. A lead generated from a static image has proven they are willing to read and comprehend the offer, making them infinitely more valuable to a retail marketplace looking for serious merchants or committed buyers.

We can see this friction play out in the reach metrics. Across the entire cohort, the median total reach sits at 2,205 users. If we look at the math, spending 23.78 Euros to reach 2,205 people yields an effective CPM of roughly 10.78 Euros for the broader sandbox. The fact that video is coming in at a fraction of that cost (0.44 Euros) proves that platforms are practically giving video inventory away to lead generation advertisers.

Yet, the smart money ignores the discount. Operators are willing to pay whatever the invisible CPM is on those static images because the back-end conversion metrics demand it.

The Hyper-Segmented Account Strategy

There is one more signal hidden in this data regarding account structure. We noted earlier that this cohort spans 63 ad accounts for just 4 brands. This means operators are hyper-segmenting their setups.

Instead of running all lead generation through a single centralized account, these brands are averaging over 15 ad accounts per brand. This structure is often used to isolate specific geographic regions, separate distinct merchant categories, or prevent different machine learning models from contaminating one another during the testing phase.

When you combine 63 ad accounts with a strict 23.78-Euro kill switch for new creatives, a clear playbook emerges:

1. Decentralize the tests: Spin up localized or category-specific ad accounts.
2. Enforce the sandbox: Cap initial creative tests (especially video) at roughly 24 Euros.
3. Fund the winners: Shift all meaningful budget (thousands of Euros) into the high-friction, high-intent static images that survive the gauntlet.

For operators running lead generation, the mandate is clear. Do not be seduced by sub-1-Euro video CPMs if the lead quality is poor. Build your own 24-Euro sandbox, aggressively cut the passive video views, and find the static images that actually drive your marketplace forward.