RAPP boosts conversion rates by 37% with cleaner customer signals

The Business Issue

Traditional marketing organizations have mastered the art of campaign delivery, with internal teams exceptionally fast and capable of building and launching sophisticated campaigns in just a few weeks.

However, efficiency does not necessarily mean effectiveness. Traditional campaign models are typically mapped out between six weeks and a year or more in advance, which means brands need to predetermine rigid, fixed customer journeys and predict what customers will want months before they actually see the content.

While this calendar-driven playbook keeps teams busy, it also creates a widening gap between brand outreach and rapidly changing customer expectations. In today’s market, consumers expect real-time, highly relevant interactions. But when brands rely on static, schedule-bound campaigns, they’re left guessing – placing the customer in the middle of a fragmented experience.

Compounding this strategic disconnect is an invisible threat: fake traffic and corrupted data signals. Globally, $71 billion was spent on fake traffic in 2021 alone, with the true downstream business impact – including contaminated CRMs, distorted analytics, and sales teams wasting energy chasing non-existent leads – projected at $204 billion.

A major study by Lunio revealed that one in 12 paid ad website clicks is fake, driven by automated web crawlers, security scanners, and bots. And because traditional campaign tools cannot distinguish human slowness from bot perfection, marketing AI models are trained using a distorted version of reality. If the data feeding a decisioning model is corrupted, the automated decisions it makes will be fundamentally flawed.

To solve this, RAPP embraced a 40-20-40 business model for AI. While traditional teams focus almost entirely on model building, RAPP dedicates 40% of its effort to getting data and security in order before selecting any technology.

Put simply, RAPP treats data as the brand’s “offensive line” – invisible when it works, but catastrophic for the business when it fails.

The Solution

To transition safely, RAPP avoided a risky “cold start.” It maintained its baseline campaigns while running Pega Customer Decision Hub™ in parallel. By implementing A/B traffic splits and a defined learning “baking period,” RAPP allowed its real-time models to learn alongside legacy content, building client confidence while protecting active revenue streams.

Surgically filtering out bad data at the point of ingestion allowed RAPP’s decisioning engine to learn from genuine human behaviors – which are naturally slow,selective, and messy – rather than the fast, perfect patterns of security bots. This dual-track approach quarantined invalid traffic for analysis while routing clean interaction data into the decisioning process.

This transformation affected RAPP’s people as much as its technology. Traditionally, creative teams spent weeks manually designing and configuring specific layouts for locked-in calendars. Transitioning to an automated, live environment allowed them to shift their role from campaign assembly to modular creative design.

Designers stopped building rigid, static emails, and instead began creating flexible content blocks. This allowed Pega’s real-time arbitration layer to dynamically assemble, test, and sequence variants on the fly. This both removed internal “creative bias” – the assumption that designers always know which visual will perform best – and allowed the machine to identify and scale hidden creative winners that traditional A/B tests missed.

The Results

The business results were immediate and significant. As RAPP expanded its use of NBA and continuously learned from customer behavior, the team reported conversion uplifts of up to 37%. Further, by cleaning two data points, RAPP achieved 30% cleaner data without changing the underlying model.

Because the system learned continuously in real time, the brand’s optimization cycles accelerated dramatically. Operationally, creative and strategy teams were liberated from manual configuration, giving valuable talent the freedom to focus on testing modular frameworks, improving treatment quality, and exploring new strategic innovations.

And by uniting clean data with real-time customer conversations, RAPP solved the universal “black box” challenge of marketing automation.

The resulting approach provided greater confidence in AI-driven decisioning by ensuring models learned from customer signals.

Want to dive deeper into RAPP’s journey? Watch their PegaWorld session here.