At Avegen, we strive to create healthcare solutions that make a positive and meaningful impact on the patients we serve. How do we know if we are creating a positive impact?

“In God we trust, all others must bring data.” – W. E. Deming

As a Data Analytics Lead, here are a few personal learnings that helped me ensure that data collection and analytics are not an afterthought.

  1. Decide on KPI and metrics
  • Every implementation with our platform solves a problem in a unique disease area. The solutions help users or patients to manage their long-term disease conditions and lead better life. There is no “one metric that fits all”. Every implementation is unique, each disease area has its own challenges. What does success look like? How do we measure success? What is the north-star metric for the implementation? It is important to discuss with your cross-functional team and external stakeholders the actionable metrics you want to collect.

2. Confirm regulatory compliance and privacy concerns

Thinking about data privacy, in the beginning, can save a lot of time and energy in the later stages of app development. When it comes to regulatory compliance, here are a  few considerations:

  • Should the data be colocated?
  • Our platform, HealthMachine™ can be extended to additional data centers. Our platform is capable of being deployed to multiple data centers across the globe which ensures that data is colocated and compliant with local regulatory requirements.
  • Providing users with the ability to turn on or off analytics
  • Are there any local laws to comply with? Evaluate supporting services during this phase. Simple things like sending a verification code via SMS will have regulations and laws to comply with in certain countries.

3. Think about data from the design phase

  • While the design team is creating prototypes and user flows, evaluate how the user will interact with the application. Thinking about analytics and events right from the design phase will help in the creation of the user funnel, help in brainstorming about user interface and design, and help in collecting user behaviour data. Ideally, the data collected must feed back in the loop to make better data-driven designs in the next iteration. Consider adding events for gestures, performance metrics and time taken to reach a defined “AHA moment”.

4. Discuss the data model and events with developers before the build phase

  • Put processes and SDLC practices in place where the cross-functional team including app engineers and quality assurance starts thinking about data and analytics when they are building a product. 
  • Decide on the data model and make sure the team is on the same page. We want to collect actionable insights, not vanity metrics.
  • Sensitize developers to make sure that they fire an analytics event upon a user interaction while the UI is getting developed. Prevent the creation of technical debt – adding the event later is a no-go. 

5. Empathize with downstream functions

  • Finally, the data collected must transform into insights. Evaluate the best visualization to tell a story. The dashboard must be minimal, easy to use and should provide its users with what they are looking for. The users here refer to external stakeholders and data controllers. Additionally, the data exports must help a statistician or AI engineer to easily process insights and model generation. Talk with their needs before creating data export pipelines.