Industries:

How PEBMED builds powerful, custom data models with Snowplow

To get the right medical information into the hands of doctors and healthcare professionals, PEBMED uses real-time behavioral data from Snowplow.

lab worker using a pipette in the lab

Background

The PEBMED app and web portal is trusted by Brazilian healthcare professionals to deliver informative, up-to-date medical content that aids in clinical decision making. 

Consisting of two subscription services, Whitebook and Nursebook, as well as a web news portal, PEBMED is the leading point-of-care mobile and web app for healthcare services. 

As PEBMED matured, they moved away from developing multiple standalone apps, to aggregating the best app content into their platforms, which are available by subscription.

Challenge

To deliver on their new business model, PEBMED needed to understand user journeys at a more granular level—which required a more mature approach to data analytics. They also wanted insights to help drive new and retain existing subscriptions, realize problem areas in the acquisition and conversion funnels, and drive decision-making in content development. 

Initially, PEBMED tracked all their products in an ad-hoc fashion, asking questions as and when they came up, and using different analytics solutions to answer them. This often led to data duplication and the need to repeatedly re-track data, as well as inconsistent naming conventions and terminology-related confusion. Furthemore, the data teams became overwhelmed with these ad-hoc requests.  

As longer-term, strategic questions arose, PEBMED began searching for a single solution to address their need for consistency and flexibility. They looked for a way to streamline and customize data tracking and collection, and centralize it to create a single source of truth. This would unlock the insight needed to tackle another challenge: understanding end-to-end user behaviors, and applying data to drive the content and product development roadmap. 

In the past we used many analytics providers and got conflicting metrics, leading to decisions based on personal opinions. Now we trust Snowplow as the single source of truth and base our decisions on those metrics, which is key for us.” 

PEDRO GEMAL LANZIERI | CTO AT PEBMED

Solution

PEBMED evaluated several third-party solutions, and found that most provided limited data ownership and transparency, applied black-box assumptions to data, and had a high latency—making the data far less reliable. 

Then, they discovered Snowplow. With our first-party data pipeline and analytics solution, they could own their data while using their own infrastructure. This gave them more flexibility through unopinionated data, plus the ability to comply with local data regulation laws, known as LGPD. 

PEBMED’s data team started off with just two data engineers, plus the support of three Brazilian startups—Datasprints, Devhat and Mobiplus—to speed up the development stage. Within a year, the data team added two data analysts, one data scientist and one data product owner. Datasprints guided PEBMED toward becoming a data-driven company. According to PEBMED’s CTO, this growth was only possible with access to the reliable data created with Snowplow. 

PEBMED’s data evolution is growing more sophisticated, thanks to Snowplow. We can see using Snowplow for everything from content recommendations to tracking internal applications and APIs to better, more granular joining up of data from individual users.” 

PEDRO GEMAL LANZIERI | CTO AT PEBMED

Why did PEBMED choose Snowplow? 

  • Own data, own infrastructure – This enables the flexibility and transparency required for deeper, richer insights, while freeing PEBMED from the huge organizational risks of black-box solutions and vendor lock-in. It also adds transparency of data management for PEBMED’s customers.
  • Flexibility to track everything – With Snowplow, PEBMED can track custom-defined events across all platforms, products and brands in a single format, ensuring data conforms to a common set of rules.
  • Centralized source of truth – The PEBMED data team can base their decisions on data and metrics drawn from custom data models and get answers to questions they would not have been able to answer before—much faster.

Results

With Snowplow, PEBMED can now pull in raw, unopinionated data from a variety of sources—which means they don’t have to fit products into a specific model or e-commerce mold. Rather, their ‘product’ is content, consumed by multiple user types that PEBMED can analyze in different ways using different assumptions. 

PEBMED is building custom data models to suit their specific needs, such as:

  • Tracking of signup
  • Content view
  • Paywall
  • Subscription
  • Page view events
  • Adding on layers of user scoring and content ranking

Their models map what users are doing in their products and with their content, and are built around a user’s lifetime value.

As PEBMED continues to build out their product offerings and develops each one to become more relevant for target users, they are thriving in their mission to integrate a data mindset across the company. Since implementing Snowplow, they have:

  • Avoided data loss due to bad tracking implementation or different output formats (i.e. application logs) in their three products. 
  • Identified the top-50 content pieces and top-5 content sections that enable subscriber retention for larger lifetime value. 
  • Created one data mart for each internal team, in order to make data accessible to all across the company. 
  • Identified user profiles that don’t engage (or, conversely over-engage) with content, and described up to six different personas for each of their products. 
  • Made data available and accessible company-wide using a visualization tool with more than 100 different dashboards representing user patterns, funnels and behaviors. 

Now, PEBMED is diving deeper into custom data modeling to make it easier for product and marketing teams to query data, and to understand more about their acquisition funnel. They are also doing more testing and experimentation with Snowplow, including personalization. 

Next steps include: 

  • Implementing data enrichment for incremental predictive analytics
  • Refining the attributes of each persona
  • Producing a mixed-content recommendation model
  • Productizing machine learning solutions to predict which customers PEBMED should invest in before and after they sign up to the platform 

How you can get started with Snowplow

To learn more about how Snowplow can empower your organization with behavioral data creation, book in a chat with our team today. Alternatively, Try Snowplow is our free, easy-to-use version of our technology, which allows you to create your own behavioral data in under 30 mins.

Ready to start creating rich, first-party data?

Image of the Snowplow app UI