The power of intent: How Visa used AI to capture high-intent users in the DRC

Milad Samia, Head of Media, Central and Eastern Europe, Middle East and Africa, Visa

As the head of media, Milad Samia drives Visa’s marketing vision for Central and Eastern Europe, Middle East and Africa. Milad has spent eighteen years at the intersection of media and marketing, driving business growth for brands and agencies.

In sub-Saharan Africa, mobile payments are often more than a convenience — they can be a critical way for people and businesses to participate in the economy. That is especially true in the Democratic Republic of Congo (DRC), where 86% of adults remain outside the formal banking system.

When Visa launched Visa Pay in the DRC, the goal was clear: help expand access to digital payments for consumers and small businesses. But reaching that goal meant solving for more than awareness. It meant finding potential users who were not just likely to install the app, but to actively start using it in their daily lives.

To do that, Visa partnered with marketing firm Assembly, to build a full-funnel campaign using Google’s AI-powered solutions. Learn how Visa used the high-intent signal of wallet creation to help identify valuable users and improve campaign efficiency over time.

The challenge: Building trust in a complex market

Launching a digital payments product in the DRC meant navigating a market with both high potential and real structural barriers. While mobile is central to daily life for many people, formal banking access remains limited. For consumers, that can mean relying on cash, fragmented payment systems, or expensive transfer methods.

For Visa, the challenge was twofold: the first was building awareness for a new product in a new market. The second was building trust.

That trust mattered because Visa Pay was designed to support real-time digital payments in everyday life. To drive adoption, the campaign needed to show consumers that the product was secure, relevant, and easy to use.

At the same time, the campaign had to grow efficiently. As with many launches, the strategy would need to evolve as more performance data became available. The risk was clear: if optimisation focused too heavily on installs alone, spend could go toward users with low, long-term value.

The insight: Intent matters more than installs

From the start, Visa looked beyond download volume as the main measure of success. Instead, they focused on a deeper signal of user intent: wallet creation.

That mattered because wallet creation required more than casual interest. Users had to complete the registration process, including document and biometric capture. In other words, this was not just an install — it was a meaningful action that showed a stronger level of commitment.

By using wallet creation as the key conversion signal, the team gave the campaign a clearer definition of what a valuable user looked like.

The approach: Using automation to learn what high-value users look like

Assembly built a campaign designed to improve as the system learned what a high-value app user was.

The first step was to launch App Campaigns for Installs. This helped generate a critical mass of data and gave the algorithm a starting point for identifying likely converters.

Once that initial learning phase was in place, the team optimised around wallet creation. Using AI, the campaigns used this high-intent signal to distinguish between users who simply installed the app and users who were more likely to complete onboarding.

The next step was to transition to App Campaigns for Engagement. This allowed the campaign to focus more deliberately on users with a higher likelihood of registering and using the app.

Branded Search and Display also played an important supporting role. Together, they helped build early visibility and capture demand as users moved from awareness into consideration.

The results: Better efficiency, stronger conversions

The shift from install-led to action-based optimisation delivered a measurable improvement in performance.

Over the first seven weeks, the campaign achieved:

What Visa learned

The launch of Visa Pay in the DRC reinforced three lessons:

First, better signals lead to better outcomes. AI-powered campaigns are most effective when they are trained on actions that reflect real business value, not just surface-level activity.

Second, automation works best when given room to learn. In a new market, overly restrictive targeting can limit performance. Letting the system learn from high-quality conversion data can produce stronger results over time.

Third, demand creation and demand capture work together. App campaigns help drive discovery, while branded search helps capture intent when users are ready to take the next step.

The launch showed that AI can be a powerful tool for scaling growth in complex markets — especially when the strategy is built around intent.