How can we help?

Overview

Company Scoring helps determine which companies are most worth reaching out to by assigning a score to every company and ranking Audiences with the highest scoring companies at the top.

A company's score is based on:

• how similar they are to your current customers and
• whether the company and/or the prospects at the company show signs of interest (buying Signals)

NOTE: while users can edit signals, adjust scores, and upload new data, we strongly recommend connecting with your UserGems CSE before making any updates to this page.

How it works

Each company's score is the sum of values calculated from four key dimensions.

1. Company Similarity 

How much the company looks like your current customers — based on tools they use, how big they are, who they hire, etc.

• Measures how similar a company is to your existing customers.
• Based on ~600 attributes per company (e.g., tech stack, hiring patterns, tools in use).
  

Example: If 77% of your customers use Salesforce, a company that uses Salesforce will receive a significant score boost.‍

2. Company Signals

Signs that the company might be interested — like if they’re hiring, just raised funding, or show intent. You can upload your own signals too.

• Includes:
  • Intent data (e.g., funding rounds, hiring activity),
  • Relationship signals (e.g., presence of a past champion).
• Signals can be uploaded and configured by the user to assign scores per signal value.

‍
3. Prospect Similarity

How much the contacts at that company look like contacts in your current pipeline.

• Evaluates how similar a contact at a company is to contacts in prospects in your open opportunities.

4. Prospect Signals

Signals from individual contacts, like if someone used to be a customer or recently asked for a demo.

• Behavioral and contextual signals attached to contacts (e.g., job changes, demo requests, past champions).
• Can include manually uploaded or automatically detected signals.

‍Scoring Calculations

Score Components

Each factor (e.g., tech stack, hiring pattern) is assigned a point value based on how common it is among customers to also do/have this thing vs the general population of companies.

We’ll use Salesloft for this example.

• ‍Customers with this factor: 23% of our current customers have Salesloft.‍
• Ratio to non Customers: out of all companies UserGems recognizes, companies with Salesloft are 18x more likely to become your customer. 

Company Size Normalization

• Larger companies are more likely to exhibit more signals. To avoid bias, a trendline adjustment is applied:
  • Expected score is computed based on company size.
    

Example: If two companies have 10,000 points, then the company with fewer employees might rank higher.

Final Ranking

• Ranking is determined by a composite score that combines:
  
  • Company score
    
  • Score of the best prospect within the company
• Prospect score influences company ranking: if no good prospects exist, the company may not appear at the top.
• If we determine that a company is good (based on having a good company score and at least one good prospect), then we show more prospects for this company to enable multi-threading.
  

User Interaction

Editing & Customization

• Users can edit signals, adjust scores, and upload new data.
  
• However, most users are not expected to manually tune the scoring model.
• We recommend connecting with your UserGems CSE before making any updates to this page.
  

Uploading Custom Signals

• Users can upload data such as:
  
  • Fit scores from external vendors (e.g., Keyplay).
    
  • Intent reports.
• Scores can be mapped to ranges or categories (e.g., A = 1000 pts, B = 500 pts).