Anyone working in lead generation knows: scaling this kind of operation is no walk in the park. Incremental gains require constant effort, ongoing testing, and close attention to details.

And when it comes to the final moments of the journey, like forms and conversion steps, the challenge gets even harder. Any change can directly impact results, which is why many companies avoid touching these critical points out of fear of hurting conversion.

Data-driven companies, however, approach this differently. Instead of avoiding the risk, they build experimentation processes to test it in a controlled way, measure the impact, and move forward with confidence.

Mercantil Bank is a clear example of a data-driven company.

With over 80 years of history, Mercantil Bank positions itself as a financial and non-financial ecosystem focused on the 50+ audience, offering products and experiences that are part of their customers' daily lives. The institution is a Brazilian reference in solutions such as loans and INSS benefit payments, combining close-knit service with a strong digital presence, including its own app and WhatsApp-based support.

In a high-volume acquisition operation where paid campaigns are a critical growth driver, optimizing conversion to increase ROI without compromising lead quality is essential to sustaining growth.

Reducing friction without compromising lead quality

Earlier this year, Mercantil Bank identified a clear optimization opportunity on one of its main landing pages. The CTAs directed users to a form that needed to be filled out before they could connect via WhatsApp. Historically, this flow had always performed well, but it raised a concern: the amount of information required up front might be creating friction and reducing conversion rates.

The question was: could reducing the number of fields to the bare minimum and collecting additional information later increase conversion? At the same time, the CRO team had a legitimate concern: fewer fields might increase lead volume, but would that come at the cost of lead quality further down the funnel?

Testing with control and confidence

To validate the hypothesis, the team structured an experiment comparing the original flow, which led users to a WhatsApp contact, against a new variation.

In the variant called popup, the form was simplified to collect only the most essential piece of information: the user's phone number. The idea was to reduce the initial effort as much as possible while still securing a contact point for future follow-up, and to speed up the path to WhatsApp.

The variant also added relevant information about the product and what the user could expect from the loan simulation. These changes preserved the original intent of the flow while lowering the barrier for the user to move forward.

Because this was a change to a critical conversion point, the experiment was run in phases. The team started the AB test with a reduced traffic allocation, closely monitoring results and making sure there were no negative impacts on conversion. As performance held steady, they gradually increased exposure until reaching 100% of traffic.

This approach allowed the team to make decisions with greater confidence throughout the process, without having to wait long periods for validation.

More conversions, more efficiency

The new variant clearly outperformed the original and delivered consistent gains across multiple metrics:

• 5.5% increase in conversion rate (form submissions).
• 15.8% increase in qualified lead rate.
• 17.5% reduction in cost per qualified lead (CPLQ).
• 23.6% reduction in customer acquisition cost (CAC).

The results showed that it was possible not only to increase lead volume but also to improve lead quality and acquisition efficiency at the same time.

Results like this give us confidence and show how experimentation can be a real catalyst for company growth, especially when hypotheses are well-grounded, and risk is kept under control.

This experiment showed that collecting more data at the start of the journey doesn't always lead to better outcomes. By reducing friction at the entry point and making the simulation result clearer (the loan amount), Mercantil Bank increased the number of users moving through the flow without compromising lead qualification and actually improving it.

The initial hypotheses held up strongly:

1. Reducing the number of form fields to lower friction and cognitive load during completion would increase form submission volume.
2. Adding supporting copy to address objections identified in prior research would improve the qualification of leads who submitted the form.

Another key learning was in how the test itself was conducted. By taking a phased approach, the team was able to mitigate risk and validate the hypothesis with confidence, even at a sensitive stage of the journey.

This test was a great example of how CRO initiatives can drive meaningful results, whether through personalization or AB testing, without requiring a major design overhaul or heavy development effort.

The case also highlights how product, marketing, and acquisition decisions are directly connected. Small changes in the user experience can drive a significant impact on business metrics like CAC and paid campaign ROI.

Wrapping up

The Mercantil Bank case shows that optimizing critical points along the journey doesn't have to be risky.

With a structured experimentation approach, it's possible to test hypotheses in a controlled way, reduce friction, and generate real performance gains, even at traditionally sensitive stages like conversion forms.

The team was able to evolve the experience based on data, increasing both the volume and the quality of leads.

In an environment where acquisition efficiency is increasingly important, this kind of practice stops being a differentiator and becomes essential.

If you also want to make product decisions backed by data, create your free account and start testing with Croct today.