To better understand how Headset is able to provide its best-in-class services, it helps to understand our process.
1
Everything at Headset begins at the retail and dispensary level. Data is collected from participating retailers’ POS systems. Headset is currently leveraged by over two thousand retailers in 30 distinct markets.
2
Data is aggregated, anonymized and normalized (mapped to a product in our database) to provide a standard view of the market. We also combine our data with information from government agencies as well as publicly traded company reporting.
3
Robots to the rescue. Machine learning algorithms help normalize the disparate product data found in the retailer data. Machine learning automates much of the process of turning messy data into something tidy.
4
Data entry and product catalog specialists ensure quality, enter new products and associate metadata with those products into product traits
5
Aggregated retailer data is cleansed to reduce noise and skew. Inferential statistics methods are applied to the sample to ensure accuracy. We also validate with product manufacturers against their sell-in data.
6
From this sample, projections are developed and utilized by Headset’s software solutions
