By Daniel Ghinn 
Today, the health strategy consulting we provide at CREATION is guided by the collective intelligence in more than one billion online conversations among more than 1.6M healthcare professionals (HCPs) all over the world. But in 2012, when we launched our project to develop the world’s first platform to study online HCPs, we started by carrying out desk research to add knowledge to what we had already observed in the online behaviours of health stakeholders.
A key factor in CREATION Pinpoint’s effectiveness is the confidence we have that the platform is tracking only confirmed HCPs – real doctors, nurses, pharmacists and allied health professionals – who typically contribute less than 5% of all online health-related online conversation. Thus the general ‘noise’ on social media is excluded from our dataset, leaving a distilled, clean dataset and a unique picture of the relationships and engagement between HCP peers on any topic at all. With CREATION Pinpoint we have tracked emerging global disease crises, pinpointed vaccine supply issues, and predicted the future of emerging treatments. All thanks to the collective intelligence of more than a million HCPs.
Whilst today we might have the world’s largest dataset of HCP conversations, when we first tested the technology in CREATION Pinpoint we had a dataset of just 5,000 HCPs, hand-picked through our desk research. By studying patterns in the profiles, networks and behaviours of those HCPs, we developed our first set of algorithms to teach CREATION Pinpoint how to identify other profiles that were also likely to be HCPs. We then put the system to work finding and ranking those candidates.
From the start, in order to be absolutely sure that the online conversations we were tracking really were from HCPs, we built an application for our international team of Data Quality Analysts. This team works all over the world, and in multiple languages, manually reviewing the HCP candidate profiles suggested by CREATION Pinpoint’s machine learning algorithms.
Data Quality Analysts carry out whatever research is necessary to decide whether a particular candidate is an HCP or not, and they classify them accordingly. Finally, a second human review assures the quality of HCP identification.
As we continued to learn about online HCP behaviours from our data, we were able to improve the algorithms in CREATION Pinpoint to include a wider range of network behaviours, so that the computer was even more effective at finding HCPs.
Today the technology hardware available to us for artificial intelligence is 400 times more powerful than the humble system on which we launched CREATION Pinpoint. So now with more than one million decisions having been made by our Data Quality Analysts, we are using a wealth of knowledge to develop a new artificial intelligence platform that is now identifying HCP profiles with incredible speed and confidence: not only is CREATION Pinpoint able to predict likely HCPs on social media, it now knows exactly how confident it is about any decision.
With improved and faster artificial intelligence at work, every decision made by a human Data Quality Analyst serves not only to classify CREATION Pinpoint’s HCP data but to improve the technology platform’s effectiveness at finding and predicting HCP profiles.
So what next for CREATION Pinpoint? We are investing in two exciting initiatives in 2019:
1) Enhanced HCP data in emerging markets, with a special focus on China, where we are already tracking thousands of HCPs on China’s most popular social media platforms.
2) An app for HCPs to use to enhance their own collaborative learning networks by providing direct access to the powerful HCP peer insights in our platform.
To find out more, or to to leverage the collective intelligence of online HCPs with CREATION Pinpoint to transform your health strategy planning, get in touch.
