How AI and iterative learning loops are revolutionising healthcare communications

A new approach to communications development that helps get you further, faster.

Article by Hannah Mann, Founding Partner at Day One Strategy

Have you ever completed extensive qualitative and quantitative research, only to end up choosing from a bad batch of concepts? None of the concepts are great. What works in the US is only okay in Germany and needs reworking for Japan. If this sounds familiar, it might be time for a change.

Pharma communications have become a complex, multi-channel beast, and what our clients want has evolved too. Faster, iterative approaches that deliver a better ROI are now essential. No one wants to finish research with something that’s just ‘okay’ and not tailored to individual market nuances.

Fortunately, it doesn’t have to be this way anymore. With the wide array of AI and machine learning tools available, we can redefine communication research best practices.

At Day One we talk about comms testing as a series of learning loops. Multiple interactions using different methods to get you the best possible outcome, at the best price, in the shortest time frame.

These learning loops can include Large Language Models (LLM’s), specially trained GPTs, and of course, human input. This allows for continuous testing and refining. We can even use AI to ideate and shape messaging and concepts before formal testing begins, leveraging our client’s existing data with the power of small language models (sLM’s).

So where does this leave us? Imagine being able to conduct AI-powered semiotic analysis on your concepts before they even reach a healthcare professional. Or using the AI equivalent of eye tracking to determine (in minutes) the optimal visual layout of your detail aid. When testing with humans, AI-generated interview summaries can be available immediately after each session. The opportunity to iterate and continually improve your materials is immense.

And we haven’t even touched on synthetic respondents. Pre- and post-testing materials with trained synthetic personas offer another layer of precision, ensuring that your final product is not just good, but great. The days of ‘test and hope’ are gone forever.

If you’re interested in learning more about Day One’s approach and how to leverage learning loops for better communication, get in touch.