The Cheapest Incrementality Testing Tool: Stella

Your Marketing Measurement is Costing You More Than You Think

Many brands assume their marketing is working because they see a strong return on ad spend (ROAS) in their reports. But ROAS is not a true measure of impact.

ROAS only tells you how much revenue was attributed to an ad, not whether that revenue was actually caused by the ad.

This leads to common problems:

Branded search ads often show 10x ROAS, but how many of those conversions would have happened anyway?
Retargeting campaigns look highly profitable, but are they just capturing customers who were already going to buy?
Prospecting ads might have lower ROAS, but are they the key to bringing in high-value customers?

The only way to separate real impact from misleading attribution is through incrementality testing and media mix modeling.

The problem is that these tools have traditionally been expensive, with costs ranging from $5,000 to $20,000 per month. This pricing has made advanced measurement inaccessible for most brands.

That’s why we built Stella, a tool that gives brands access to real incrementality insights for a fraction of the cost.
‍

How Much Money Are You Wasting Right Now?

A performance marketing agency running Meta and Google Ads for a well-known ecommerce brand recently conducted an incrementality test. They discovered that 38% of their reported revenue was non-incremental.

Before testing, they assumed their ad campaigns were driving most of their sales because their blended ROAS looked strong. However, when they ran a geo-based holdout test, they saw that many of their sales would have happened organically.

By reallocating budget based on true incrementality, they increased their incremental revenue by 21% without increasing spend. Instead of relying on platform-reported numbers, they shifted budget away from non-incremental campaigns and saw real profit growth.

If you are not testing for incrementality, you are making budget decisions based on incomplete data.
‍

Why Traditional Marketing Measurement is Broken

If you are still using ROAS, first-click, or last-click attribution, you are only getting part of the picture.

The Myth of Branded Search Ads

Many brands see high ROAS on branded search campaigns and assume they are critical to performance.

One Stella client, an ecommerce brand in the beauty industry, was spending $15,000 per month on branded search because it showed a 9.5 ROAS.

When they ran an incrementality test, they found that 81% of their branded search conversions would have happened anyway through organic search or direct traffic.

They reduced branded search spending by 75% and saw no significant drop in revenue. Instead, they reallocated that budget to prospecting, which increased new customer acquisition by 18% over the following months.

If you have never tested the incrementality of branded search, you could be overspending on ads that are simply capturing demand that already exists.

Testing branded search also tells you exactly how incremental it is so you can adjust your budgets accordingly.
‍

Causal Impact Analysis with Stella

If you searched for terms like “causal impact analysis” or “data impact analysis tools,” you might be wondering whether Stella can perform these types of tests.

Yes, Stella supports causal impact analysis, but with more flexibility than traditional models like Google’s CausalImpact package.
‍

How Stella’s Causal Impact Analysis Works

GeoLift-Style Testing – Stella’s primary model uses a geo-based holdout approach, which is more reliable for measuring ad impact at scale.
Causal Impact Mode – If you want a time-series-based causal impact analysis, you can switch models in Stella’s “Advanced Options” section of the final analysis tab.
‍

Why marketers use Stella instead of Google’s CausalImpact

Automated holdout selection, so you do not have to manually match geographies
Faster results, with tests requiring as little as 20 days instead of 45+
AI-driven recommendations, so you get clear next steps rather than just raw data

For brands looking for an alternative to open-source causal impact tools, Stella provides a more robust solution designed specifically for marketing measurement.
‍

Why Stella is the Only Tool That Makes Incrementality Testing Accessible

Most incrementality platforms are built for enterprise brands with large budgets. Their pricing and complexity exclude mid-market brands that actually need these insights the most.

Stella changes that:

‍

Unlike other platforms that require a $50,000 to $200,000 holdout budget, Stella is designed for efficient testing that requires as little as $5,000 to $15,000.
‍