Companies Classification Methodology

Emsi Burning Glass has replaced its previous company normalizer with a new companies classifier. The previous company classification had become outdated and difficult to update. Our new company classifier is built from over 400 different open source company datasets and draws from each to provide a comprehensive view of companies in the world. This classification is designed to be global in scale and adaptable to the future needs of our customers.

Why replacement is important: 

1. Global in scale:  We can use the company classifier to classify job postings and profiles in all of the countries for which we have data.
2. Recency: The Emsi Burning Glass Companies Taxonomy is updated every two weeks to keep pace with shifts in the market. 
3. Adaptable to fit customer needs: We can develop and add new metadata to our Companies Taxonomy to better serve our clients’ evolving needs. 

How it works: 

Starting with raw company names, we normalize these names using a set of proprietary criteria. This strips information from the name that is irrelevant to identifying the company correctly (e.g. LLC, Inc.). This leaves a normalized name with all the ingredients needed to classify it to our Companies Taxonomy.

After normalization, we match the clean name to the best fit in our Companies Taxonomy. Each company has associated metadata, including Tradestyle, NAICS codes, and staffing labels. If a company is a subsidiary or establishment of another company, we generally roll it up into the main company when the establishment or subsidiary has the parent in its name. For example, “Walmart Canada” would be classified as “Walmart.”

NAICS methodology:

We code companies to NAICS at the 6-digit level where possible, and will fall back to 2-digit NAICS if we are unable to infer a 6-digit NAICS. While we realize a company can work in multiple industries depending on the establishment, currently, we assign only one NAICS code to each company. This is based on the most common industry a company works in at the establishment level. For example, Amazon is coded to 454110 – Electronic Shopping and Mail-Order Houses, even though many Amazon establishments may work in different industries.

We currently use US NAICS for both US and Canada. Canada has their own NAICS code system that differs from the US. However, we have not added this industry taxonomy to Canadian companies yet.

Staffing Methodology:

Companies are labeled as a staffing company based on name, industry code, and qualitative research. For the purposes of job posting data, companies are labeled as staffing when they are a) true staffing companies, or b) job boards or brands maintained by staffing companies. This allows customers to filter results based on what they would like to see.