Takeda's NBA AI boosts new drug uptake about 2.35 times

Takeda's go-to-market overhaul responds to a shift it describes in how medical professionals gather information: their sources have multiplied, and no single, uniform pitch delivers enough value anymore. Rather than simply adding digital channels, Takeda says it is combining its medical representatives' activities, digital channels and AI analysis to tailor engagement to each individual. The company says it is testing new commercial models on five products with different characteristics in fiscal 2026, and plans to spread what it learns to other products in Japan and to medical-professional experiences globally.

The NBA recommendation engine sits at the center of that effort. Takeda says it draws on delivery records, digital behavior data, Real World Data and market data, and that recommendations have both improved its understanding of medical professionals and led to business results. Takeda attributes these gains not only to the data and AI itself, but to an operating setup that keeps improving the models: a shared global data and AI foundation adapted for Japan, in-house specialists who refine models for seasonal vaccines or rare-disease areas with limited past data, and continuous dialogue between developers and field users.

Parallel work targets content. Takeda says it has brought AI into planning, production, quality control and use, learning from past review comments and approved materials to catch missed corrections and updates. That has shortened the path from planning to publication from several months to as little as about one month, while structured content feeds the NBA system. The Experience Core guidelines and the Takeda ID user figures suggest Takeda is also trying to measure this value. Whether the early uplift holds across products with less data, and whether these gains extend beyond Japan, is likely to shape how far the transformation spreads.