Fixing SIC codes
By Bernard Lam on the 18th November 2022Industry Artificial Intelligence
Competition drives innovation, and understanding how prospects stack up against their competitors is crucial, whether you're in private equity, finance or business development. The most commonly used method for company comparison is through Standard Industrial Classification codes (SIC Code), but there’s a problem. These classifications are dated, leading to poor data, invisible peers and missed opportunities. Many of the economic metrics we use today were developed during the industrial age by big companies like General Electric and ExxonMobil, with physical plants and a large output of material goods. There were no tech companies back then, at least not as we currently define them.
The UK SIC was last revised in 2007, and out of a total of 992 SIC codes, there are roughly 250 codes for companies involved in Manufacturing. However, nearly a third of all UK companies are classified under one of the 74 SIC Codes that start with ‘Other’. Some of the most valuable companies in the world: Apple, Alphabet (Google), Amazon and Microsoft. These businesses generate revenue in various ways — Apple from hardware, Microsoft from software, and Google through advertising, and though they do share some similarities, IT doesn’t seem like the right category to group all of them into.
With all the amazing technology at our fingertips, we believe there’s a better way of doing it. Instead of focusing on vertical industries, it’s time to look at business models. At Scribe, we’ve begun pushing innovation, training a Natural Language Understanding model to define companies by what they publicly say about their own activity. This means that when you search for “payroll software”, you’ll get payroll software companies.
Scribe is an API-first company data platform, powered by research-based AI. Uncover deep insights about private companies and end data entry from PDFs.
Header Image by manuelsechi