Middesk's Industry Classification (IC) product is an enterprise-grade, scalable solution for discovering what every business makes and sells. With a single request, Industry Classification retrieves the web presence of a business entity, extracts and analyzes its content, and leverages a proprietary machine learning model to determine the industry of a given business.

On top of flagging businesses operating in prohibited/risky industries, Middesk's Industry Classification provides full coverage across all of North American Industry Classification System (NAICS) codes at the 4-digit level, and equivalent Standard Industrial Classification (SIC) codes.

  • Note: we are able to classify businesses across all 4-digit NAICS codes, but the confidence of the response is dependent on the inputs that the model receives.

There are two versions of Industry Classification for customers to choose from:

Product Flow

At a high level, the product is designed as follows:

  1. A customer passes Middesk a business website either through the Middesk Dashboard, Business API or Industry-only Order API.
  2. Upon receiving the website, Middesk aggregates the web content associated with that website, collecting information across numerous pages of the site.
  3. Middesk sends the aggregated web content to a proprietary machine learning model to develop a set of classification label predictions.
  4. Middesk returns this full set of label predictions to the customer through the Dashboard or via webhook.
    • The output is dependent on the version of IC that the customer has opted into.


What’s Next

Check out the detail list of categories (both High-Risk and Premium) in the following page.