Effective marketing hinges on identifying and speaking meaningfully to nuance in human experiences. While many marketing verticals can use data like audience engagement metrics to capture that level of nuance, directly informing sales strategies and campaign optimizations, regulation rightfully limits that ability for pharma. It’s intriguing that an industry aiming to reach audiences during such deeply personal and emotional journeys is still looking to automotive and CPG for inspiration. It’s even more intriguing to see investments in AI going toward more data to support further fragmentation of familiar strategies… It feels like we’re paying to see a film in IMAX 70mm and then using that screen to watch a standard format movie instead.
AI comes down to mathematical representations of human expression. LLMs use math to construct sentences and words have assigned values so they can be understood by computers. This “map” is called an ontology, and it requires upkeep as meanings change. This means we can measure the distance between two words, and as language changes through semantic drift, we can roughly measure change in meaning.
The concept of semantic drift isn’t new, nor is measuring it. What’s new is how pharma marketers can benefit from it by comparing drift to approved messaging in market. In this capacity, we can use AI to support the creative process instead of poorly replacing it. Measurement of semantic drift could become the closest approximation of change in audience thinking that traditional media metrics currently fail to capture. It also ties creative and media work more closely together, making space for conversations around message strategy as a baseline for media measurement, something social listening and qualitative brand impact surveys already aim to capture. This is harder to do for DTC audiences than for HCP, of course, but it’s not impossible. Even with strict compliance protocols in place, scrubbed and aggregated insights can allow pharma to drop rationalizing “what worked” and dig more into thinking about “why.”
Written by Dr. Nadia Khatri, unassisted by generative AI
Published September 15, 2026