From predictive sentiment analysis and word association to audience profiling and message personalization, social listening techniques are helping healthcare marketers translate everyday conversations into brand positioning strategies, outreach programs, and relevant online content.
With the exponential growth in social sharing and social media, we posed the question, “Are You Listening” to the healthcare industry during a recent panel discussion on social listening at Ogilvy CommonHealth Worldwide’s 3rd Annual Marketing Analytics & Consulting Summit. The reaction to the discussion during the summit was incredible, as attendees bombarded our panelists with questions, which made for a lively discussion.
Joining our expert panel discussion were several contributors: Ryan Alovis, InTouchMD, Karen Auteri, IMS Health, Michele Baer, Feinstein Kean Healthcare, Kim-Fredrick Schneider, Sermo, and our very own Angelo Campano, Ogilvy Healthworld.
Attendees learned multiple perspectives from our expert panelists. First, social listening provides marketers with a reality check for what patients and physicians are discussing in terms of disease states, available drugs, and lifestyle considerations. Second, attendees learned that many of the techniques employed have been shown to help marketers manage and respond to adverse events and reposition web content to deliver more meaningful messages to audiences they are trying to reach and educate.
Making sense of social conversations as related to branded and unbranded messages, and disease states, is central to capturing emerging patient and physician trends around sentiment, preference, and message personalization. In the Analytics department at Ogilvy CommonHealth Worldwide, we believe social listening needs to be a dynamic discipline that is “always on” and can be configured to leverage our sophisticated network of algorithms to aggregate unstructured conversations, and glean meaningful insights related to the way patients and physicians are talking about our clients’ products.
Natural Language Processing (NPL) and text mining machine learning algorithms are used to extract dominant concepts across posts, tweets, text messages, and call center conversations. We create a dictionary of terms with the highest frequency across messages, which is also known as a term document matrix. Correlation analyses are run across the document matrix to isolate the top 100 concepts and messages. This concept investigation is done through splitting the data into a training dataset and a test dataset (usually a 70/30 split, respectively). We then apply decision trees and neural networks to learn from our sample training data on how the text in each comment is configured to help derive classification rules on sentiment (positive or negative). Once classification rules are set, our rules are then deployed for overall monthly scoring of brand sentiment.
We can help our clients understand questions such as:
- What are HCP and patient sentiments about the brand?
- What are the terms and attributes HCPs and patients are using to refer to our ailment state or specific brand?
- What are HCPs and patients saying about competitor brands?
- How can we proactively manage adverse events reporting?
With limited social buzz, a cancer drug maker found that their brand’s category was mostly associated with terms like LDK-378, crizotinib and maintenance terms. The brand itself was strongly associated with terms like Tarceva and ALK, but social listening allowed the brand to identify opportunities within the category to purchase tertiary or long-tail terms to optimize search.
In addition to finding ways to optimize search, we were able to identify three different types of back pain sufferers through social listening. From over 115,000 local EU market conversations, we were able to identify pre-concerned, seekers, and diagnosed back pain sufferers. This learning enabled our marketing plan to amplify key brand messages at the right moment, in the right space, and at the right time that was most relevant to when each audience was most likely to respond.
Offering Many Benefits
Through understanding and evaluating the reality of how patients and physicians are talking about disease states, branded or unbranded products, we’ve reshaped website content, fine-tuned campaign messages, optimized SEO, and considered new targeting pathways. Our processes will continue to evolve to help drug manufacturers become more relevant in meeting physician information and patient care needs.
If you’re not listening, our Analytics group at Ogilvy CommonHealth Worldwide can help get you started.
CONTINUE THE CONVERSATION:
Questions? Comments? You can contact the author directly at firstname.lastname@example.org.
Please allow 24 hours for response.