Pharma content personalization engine delivered in 4 weeks

ClientTop 5 Pharma Company

Location

Germany flagGermany

IndustryPharmaceutical Manufacturing

About the project

We helped a leading pharma company personalize their HCP portal based on physician specialty

The client, a global leader in pharmaceutical manufacturing, runs a login-based HCP engagement portal serving physicians across multiple specialties. The team wanted to increase the number of premium subscribers and increase engagement with the platform.

We delivered the End-to-End Deployment & Launch tier of our HCP Personalization Engine service.

The Challenges

The client was navigating the reality of personalizing an established HCP portal: content that wasn’t centralized and labeled, no clear performance metrics in place to see how physicians are engaging with it.

Because the portal served sensitive clinical audiences, any new personalization logic also had to hold up to close compliance scrutiny around how HCP data was used.

Physicians, for their part, tended to engage only with content pushed through campaigns or sales outreach, rarely exploring the rest of the portal on their own.

The Objectives

With the platform redesign underway, Digitalya and the client aligned on a focused set of goals: raise the number of premium subscribers and increase engagement with the portal.

01

Define multiple personalization strategies and algorithms for the portal from the ground up.

02

Design a model the client could launch quickly despite untagged, uncategorized content.

03

Use HCP specialty data already captured at login as the personalization signal.

04

Grow adoption of the premium content tier.

05

Extend personalization from the homepage into individual article pages.

06

Deliver the solution within a short, sprint-based timeline.

Our Approach & Solutions

Digitalya built the personalization engine around data the portal already collected: HCP specialty at login. The team presented two implementation paths and worked with the client to choose the one that could launch the fastest. The next step was carrying that same logic from the homepage into the individual article experience.

  • 01

    Login-based HCP specialty profiling in place of a tagging-dependent model

    Since the portal already required a login, the team used each HCP’s registered specialty as the core signal for personalization.

  • 02

    Two algorithm options presented; a basic specialty-based model chosen for faster launch

    Digitalya proposed a specialty-based model alongside a more advanced option that would have grouped recommendations by reading history and bookmarked articles. The client chose the first option to prioritize a fast, low-friction launch.

  • 03

    Dynamic homepage batches of six personalized articles

    A homepage module was built that displayed six personalized articles at a time, matched to the visiting HCP’s specialty. HCPs can load additional batches as they continue to browse.

  • 04

    Premium articles surfaced within recommendations to drive upgrades

    Each page batch deliberately includes premium articles, placing that content in front of the HCPs most likely to value it. The goal is to encourage more physicians to upgrade their accounts.

  • 05

    Sidebar recommendations of up to five specialty-matched articles on every article page

    HCPs tended to engage only with content pushed through campaigns or sales reps, rarely exploring further on their own. The team addressed this with a sidebar module on individual article pages that could surface up to five additional specialty-matched articles.

  • 06

    Back-end adaptation to handle inconsistent physician data from Veeva

    HCP data feeding the personalization engine was pulled from Veeva, which introduced integration complexities as the project came together.

  • 07

    Consultative guidance on performance metrics for measuring future campaigns

    Digitalya proactively recommended a set of performance indicators the client could adopt going forward, extending its role beyond deployment and launch into strategic advice on measurement.

The Results

The personalization engine went live faster than the client expected, and its impact reached beyond the numbers. Just as notable was the shift in the relationship itself: the client came away seeing Digitalya as a partner it could trust to shape strategy, not only execute against a brief.

Key outcomes included:

  • 01

    Delivered in roughly four weeks, 50% faster than the client expected

  • 02

    Estimated 10-20% in premium subscriptions

  • 03

    Sidebar recommendations shifted engagement beyond campaign-only entry points

  • 04

    A basic personalization model the client can build on as content tagging and performance metrics improve

Highlights

4 weeks

from concept to launch

10-20%

estimated increase in premium subscriptions

5

personalized recommendations per article

Technologies
Vue.js
TypeScript
Image showing the PHP logo
PHP
Java
Image showing the Veeva logo
Veeva
PostgreSQL
PostgreSQL

What stood out on this project was the kind of partner the client would lean on. They could come to us and say “I want something, but I’m not sure exactly what” and we’d step in, propose a couple of directions, show them examples from other work, and ask them what would fit them best. Being trusted to work that way, instead of just executing a fixed brief, is what made this one memorable for us.

Daniel Brunescu

Daniel Brunescu

Delivery Manager, Digitalya

What stood out on this project was the kind of partner the client would lean on. They could come to us and say “I want something, but I’m not sure exactly what” and we’d step in, propose a couple of directions, show them examples from other work, and ask them what would fit them best. Being trusted to work that way, instead of just executing a fixed brief, is what made this one memorable for us.

Daniel Brunescu

Daniel Brunescu

Delivery Manager, Digitalya

Need help with a complex content migration?

We can help you audit, structure, migrate, and validate your content faster.