Case Study: How Measurement Made the Difference in X Ad Testing

Testing new marketing channels can feel like a gamble. You don’t have the luxury of wasting budget on platforms that don’t perform—and with complex user journeys, it can be hard to know what’s truly working. That uncertainty can keep you from exploring growth opportunities.

In this post, we’ll share how we helped a SaaS client overcome that exact challenge—using a strategic, low-risk testing approach to uncover the real value of a new ad network, unlocking efficient lead generation.

The Challenge

Our SaaS client wanted to scale lead volume while hitting strict efficiency targets. We identified the social network X as a promising expansion opportunity, but based on the way users interact with the platform we anticipated significant out-of-channel (OOC) impact—where users are influenced by ads but don’t click on them directly. This posed a measurement challenge. If the majority of conversions aren’t directly trackable, it becomes difficult to accurately value the network and set appropriate bid targets. To run an effective campaign on X, we needed a reliable way to quantify its true impact, including OOC conversions.

Our Approach

Measuring Lift

We used location data to launch a geo test targeting two large U.S. regions while holding out other states as control groups. We assessed regional lead volume before launching the test to understand how volume trends compared between our identified test and control regions. After launching paid ads in X, we continued to measure regional lead volume and look for changes in the trends. With the launch of X campaigns being the only regional treatment applied, we could attribute increased lead volume in the test region to the new campaign, regardless of which source our attribution model identified. 

Test Structure

We designed the test with two key objectives in mind:

1. Measure the Real Impact of X 

   We aimed to estimate the OOC lift with enough confidence to inform future spend decisions. While no OOC factor will be perfectly precise, our goal was to generate reliable directional data to guide efficient scaling.

   
   

2. Minimize Risk

   During the test period, we expected to "overpay" for leads—because we hadn’t yet validated the actual value driven by X. We deliberately kept test spend limited, balancing data needs with budget efficiency.

With these goals in mind, we ran the test for 4 weeks. This allowed us to conclude the test as quickly as possible while still giving the campaign enough time to ramp up and generate results for a few weeks, increasing our confidence in the validity of the outcomes. 

The Results

As expected, directly tracked leads from X were low. But the OOC geo test provided deeper insight: for every lead directly attributed to X, the test suggested there were 18 additional leads influenced by the campaign.

Equally important was understanding where this lift showed up. If most of it flowed into branded search, for example, that would impact how we value and bid on branded keywords as we expanded into X. In this case, most of the lift came through direct and organic channels. Armed with that insight, we confidently expanded the campaign nationwide. Since launch, X has helped drive a nearly 30% increase in organic lead volume. 

Without the OOC test, a campaign in X would have struggled to meet the client's strict efficiency benchmarks—and we would have missed out on a high-performing channel.

Conclusion: Why Measurement Makes A Difference in Ad Testing

Testing new ad networks doesn’t have to be a shot in the dark. With the right approach, you can reduce risk, uncover hidden value, and make smarter investment decisions—even when direct attribution falls short. For our SaaS client, a thoughtful geo-based test revealed the true impact of a high-potential channel, unlocking meaningful lead growth while staying within strict efficiency goals. If you're facing uncertainty about where to invest next, remember: the key isn't avoiding risk—it's testing in a way that gives you confidence to scale.