Unilever cuts product development time from months to days using AI
Unilever's Beauty & Wellbeing division is using AI to compress product development cycles that once took five or six formulation rounds into one or two. The company now moves from identifying a consumer trend to launching a science-backed product in days instead of months.
The shift stems from a straightforward problem: beauty shoppers have changed. Eighty-seven percent now discover products via social platforms, and 91% of US adults aged 30-54 say they're more ingredient-aware than before. Trends move fast. Product development had to follow.
Mining a century of research data
Unilever's scientists now use AI to assess consumer insights 60% faster than previously possible. The company analyzes over 1,000 external data sources each month-social media, search queries, retail activity, competitor moves-to surface what's actually driving conversations about beauty and wellbeing.
Once a trend emerges, teams feed that insight into proprietary AI tools connected to Unilever's R&D databases. This means researchers can quickly search decades of ingredient libraries, formulation trials, sensory tests, and consumer studies to design products that match what shoppers actually want.
The result: concept-to-R&D-brief time dropped from months to days. Claims generation is 75% faster.
Virtual testing cuts research time and cost
Unilever built structured datasets from its microbiome research to power what the company calls "virtual cohorts"-AI-generated sample groups that represent specific demographics based on age, skin type, hair type, and location.
Scientists can now test how around 2,500 virtual subjects would respond to formulas, claims, and sensory profiles simultaneously. This approach doesn't replace real-world testing but lets researchers evaluate concepts faster and cheaper before physical trials begin.
The company also deployed an internal AI assistant that connects 150,000 scientific documents spanning over a century of research. Scientists query the system in natural language to surface insights that would take weeks to find manually.
Real products, real results
Two recent launches show what the process produces. Pond's Skin Institute's Hydra Miracle range features Cera-Hyamino™ technology-a blend of ceramides, hyaluronic boosters, and amino acids designed to strengthen skin barriers. The ingredient delivered clinical results showing 78% more hydrated skin from day one. AI analysis of Unilever's microbiome data identified the ingredient combination faster than traditional research methods could.
Dove's Damage Therapy range uses Bio-Protein Care technology to replenish hair protein lost to heat styling and coloring. Scientists at Unilever's Materials Innovation Factory analyzed over 100,000 data points on hair properties using advanced measurement tools and robotics. The company filed five patents for the active ingredients and reported double-digit growth for Dove in 2025, with the Damage Therapy range rolling out to over 35 countries in the third quarter.
Jason Harcup, Chief R&D Officer of Unilever Beauty & Wellbeing, said the shift goes beyond speed. "For our 4,500 researchers, AI isn't just a time-saver. It's changing how we discover, collaborate and innovate."
What this means for product development teams
AI for Product Development now means faster trend detection and shorter cycles between concept and launch. Teams can leverage AI Data Analysis to process historical research and consumer data that would overwhelm traditional methods.
The shift doesn't eliminate human expertise. It frees researchers from manual data work so they can focus on interpretation, strategy, and the creative problem-solving that AI can't do alone.
Your membership also unlocks:
