The Customer
A Switzerland-based drugmaker wanted to mine data on HCP activity across Twitter and LinkedIn to generate valuable insights that can help its sales representatives personalize and nurture relationships on these digital platforms.
Challenges
- Difficulty managing the high volume of diverse HCP data on social media
- Unstructured and fragmented data created challenges in identifying the right target audience
- Limited insights to design laser-focused marketing campaigns
The Solution
We built an end-to-end Natural Language Processing (NLP) and Machine Learning (ML) model to structure unstructured HCP social media data, capture information units such as organization, location, medication, etc., and generate insights into HCP opinions and sentiment on a particular topic. We leveraged data mining techniques and technologies such as SciSpacy, BioBERT, BERT, Google Encoder, and AWS Comprehend. To learn more about how we executed this strategy, download the PDF below.
Outcomes
An NLP-driven approach to analyzing social media data helped the customer shed light on key insights such as types of disease-related discussions trending among HCPs, common topics discussed within a specific period, and more.
12%
Social engagement
60%
Speed to insights
70%
Time to persona building