Have you ever wished you could know in advance which physicians were about to switch away from your brand, and what you need to do to retain them before they switch? Do you wish you knew which physicians were most important to keep (not only in terms of prescribing but also in terms of influence and complete value of that customer)?
These insights are critical to companies with larger brands that may be losing exclusivity or facing a competitive threat. Sales reps from all competing companies see the same physicians. Once a company has convinced a physician of the clinical superiority of their brand, how do they know when the competitor is starting to gain traction and convincing the physician to leave the brand for their own? No matter how hard a company tries, competitor sales reps will be seeing the same physicians as their reps, providing persuasive arguments as to why they should switch from the client brand of drug to the competitor one. Once the client has invested so much time and money convincing a physician about the clinical merits of their brand, how can they ensure that a competitor does not entice the physician to switch to their competing brand?
Many factors influence brand choice in Pharma and, of course, Pharma sales and marketing influences that choice. Identifying at-risk physicians before they leave your brand is especially critical to many larger brands and also brands that have lost exclusivity. Companies have spent a lot of money on sales and marketing to gain that customer in the first place, but the level of switching and also the cost of approaches to retain them can be high. Companies need to know how to retain them before they switch, and who is most important to retain, as well as what the most cost-effective and yet effective means is to retain that physician. For some physicians it will not be a sales call but an email with an appropriate journal article attached, or something similar.
AI Powered Analytics to Predict
Using AI powered analytics, we can tell what channel and content will provide the optimal result for an individual physician… and it may not be a sales call. The cost of switch within the Pharmaceutical industry can be extremely high. Think about Lipitor and what happened when the patent expired; given how large a brand it was, the profits of Pfizer as a whole declined by half when it went off patent. Lipitor was a much-loved brand and the decline in revenue was in the billions. Not all of those that switched to the generic would have if they had been identified in advance, and communicated to appropriately to keep them. What will happen to Abbvie when Humira faces a similar challenge?
On the flip side, we can also use AI powered analytics to predict if a physician who is on a competing brand is starting to consider switching to your brand, which would be of particular interest for smaller brands. Predicting which physicians are starting to consider switching to your brand and communicating the right messages to those physicians, in the right channels, would enable the switch more rapidly. When analyzing physician behaviour, we can equally predict switches from as well as switches to brands. There have been older models which have had their place; they might have examined this behaviour, identified what factors they believed would indicate this change in behaviour, and when people exhibited those behaviours they picked them up. However, there is a much larger range of behaviours that physicians exhibit that you may not be able to predict.
Eularis use AI to do this far more reliably. We do not pre-choose the behavioural factors ourselves but use the power of big data and AI to correctly identify the right combination of factors, as well as the temporal sequence of the associated behaviours, to make reliable switch predictions that capture over 90% of the switchers in advance. We can then ensure that they stay with our brand (or move to our brand) with targeted and appropriate communications based on their needs. Eularis can also identify why the switch has happened, which can lead to considering other interventions in order to avoid the situation in the first place. By using this approach, and having the data constantly being refreshed and learning from the results, we can achieve 100% accuracy on the training dataset, and in the high 90% range with the actual dynamic datasets.
In order to be able to do this, we need data. Eularis identify all sources of data available in that market and company, and then integrate the data sources to enable us to track what actions the physician is taking (e.g. what websites the doctor is going to, what articles they are reading, what they are saying to the sales reps, etc.). We are also able to track what actions lead to what prescribing outcomes for individual physicians. By combining this with other data, we can write AI code to understand what physician behaviours are warning signs that the physician is planning to switch to (or from) a competitor. Typically, once we set up the system, we would need 1-2 months of data to start getting reliable feeds. We can then determine what channel and content is best to keep or switch the physician to our brand. If the channel is a sales rep, the appropriate sales rep could get an alert each day of the physicians needing to be seen in addition to what content should be communicated to each individual physician to influence them the most. In this case, the sales rep for the relevant physician would get an alert each day of any physicians flagged, then can use the AI generated content suggestions and go to see that physician to convince them to stay with our drug before it is too late to change his/her mind.
There are, of course, many challenges in implementing this approach, including combining many disparate data sources and platforms, implementing appropriate tracking tools, applying the right AI techniques, and ensuring the right sales and marketing teams get the data they need in time to influence that physician to change their behaviour. Another obstacle is ensuring that the teams follow the information that the AI is creating and create a process to respond to the information in a timely way in order to mitigate the results. The reason to switch is not from one factor but multiple influences, and without using machine learning techniques (and especially deep learning) to make sense of these and learn, this high level of accuracy cannot be replicated. Speed to getting to that physician is very important as once they move, they may not switch back if they are happy with the competitor brand.
