Personalization & Hyper-personalization Engine

This project started with a simple observation. We had enough traffic and enough communication channels, but we were speaking to everyone in the same way.

The objective was to replace broad audience campaigns with behavior driven experiences that matched each user's interests, activity, and stage in their investment journey. I worked on audience segmentation, campaign strategy, content planning, and personalization across multiple customer touchpoints. While building this capability, I also led a detailed evaluation of enterprise personalization platforms to understand what infrastructure would support the company's future growth.

Financial products compete for attention every day. Users receive countless notifications, emails, and investment recommendations from different platforms. Generic communication quickly becomes background noise.

The challenge was not sending more campaigns. The challenge was making every interaction feel relevant. Better personalization promised higher engagement, stronger customer relationships, and more efficient use of every marketing channel.

I approached the project as a long term capability rather than a collection of campaigns. The goal was to build a framework that could continuously learn from user behavior and improve every future interaction.

When I reviewed campaign performance, I noticed that low click through rates were not caused by poor creatives alone. The larger issue was that every audience received nearly identical communication regardless of their interests or investing experience.

Someone researching SIPs, a first time investor exploring finance basics, and an active trader could all receive the same notification. Even well designed campaigns struggle when relevance is missing.

Before discussing personalization, I wanted to understand who our users were, what they cared about, and how those interests changed over time.

I avoided creating segments based on a single attribute. Instead, I combined multiple signals so every audience represented meaningful differences in behaviour.

Personalization extended beyond notifications. The same behavioural understanding influenced banners, popups, articles, email campaigns, SMS, WhatsApp communication, and in app experiences.

Every interaction became an opportunity to learn more about user intent. Better understanding produced better recommendations, and better recommendations encouraged stronger engagement. Over time this created a continuous improvement cycle instead of isolated campaigns.

Personalization worked only when it remained consistent across every channel. A user who showed interest in mutual funds on the website should not receive generic trading notifications on mobile the next day.

I planned campaigns so that each channel played a different role. Website banners captured attention. Push notifications encouraged return visits. Email delivered richer educational content. WhatsApp supported timely communication while SMS was reserved for important updates where immediacy mattered.

Better targeting alone was not enough. The content also needed to reflect what users were trying to accomplish. I worked with design and content teams to align articles, creatives, and landing experiences with observed behaviour instead of publishing the same message for every audience.

As personalization became more sophisticated, I wanted to understand whether the existing technology stack could support future business goals. I independently evaluated leading personalization platforms to compare their strengths, limitations, integration effort, reporting capabilities, scalability, and long term suitability.

The assessment covered products including CleverTap, Mixpanel, MoEngage, WebEngage, Adobe, and emerging AI focused platforms. The findings were presented to leadership to support future platform decisions based on business needs rather than feature lists.

Building the framework required much more than creating audience segments. Every campaign needed clear entry rules, content mappings, frequency limits, and success metrics before it was launched.

I worked with product, design, content, engineering, and marketing teams to make sure every experience remained consistent across channels. Small improvements were released continuously so campaign performance could improve over time instead of waiting for one large launch.

Every campaign was evaluated using engagement and conversion metrics rather than delivery numbers alone. This helped distinguish messages that reached users from messages that genuinely influenced behaviour.

As audience targeting became more precise, campaign performance improved steadily. Notification click through rate increased from 0.2 percent to 3 percent, demonstrating the value of sending fewer but more relevant messages.

The initiative also established a repeatable personalization model that could support future product launches, customer journeys, and marketing campaigns without rebuilding the strategy from scratch.

This project changed how I think about personalization. Good personalization is not about showing different banners to different people. It is about understanding why someone came to the platform and helping them move forward with confidence.

Looking back, the biggest lesson was that better technology alone does not create better customer experiences. Clear strategy, thoughtful segmentation, and continuous learning are what make personalization valuable.

Today I would combine behavioral segmentation with real time AI models, predictive recommendations, and experimentation that adapts content during a user's active session. I would also introduce a centralized decision engine so every customer touchpoint could use the same personalization logic.