How it works

Cosmention is based on cutting-edge research in natural language processing (NLP). It uses deep learning to analyze text in social media posts. We have trained custom models specific to cosmetics and enhance their usefulness with data on product listings in online shops.

Every day, Cosmention updates its data stories in 3 steps:

Illustration of social media data collection

Social media data collection

Cosmention collects new posts from a curated panel of trending Blogs, YouTube channels, Instagram and Twitter hashtags and accounts, as well as Reddit communities. In addition, we use web search query frequency data from Google, product listings in online shops as well as reviews.

Illustration of data processing

Natural language processing

We use cutting edge natural language processing through deep learning. We have custom trained models optimized for identifying mentions of beauty brands and products in social media posts. Our models also perform time series forecasts and cluster analysis.

Illustration of data reporting
Illustration: Veronika Simmering

Data storytelling

Cosmention compares the new data to historical data and predictions. If a product, brand, ingredient, hashtag performs differently than before, it automatically creates a data story for your news feed. It also reports on new products and brands, as well as consumer sentiment. All data and analysis is also available in a dashboard.