AI Outbound Scoring

AI Outbound Scoring uses AI to build a configurable and transparent scoring model that ranks accounts across HG Insights' full universe of 50 million+ companies,  by Fit, Need, and Intent, surfacing the best accounts for your team to go after before they ever land in your CRM.

Unlike scoring approaches that only work with accounts already sitting in your CRM, AI Outbound Scoring works across HG's entire account database. That means it can surface strong-fit accounts your team has never worked before, not just re-rank the ones you already have.

AI Outbound Scoring is included in your RGI Platform subscription with any module — Market Analyzer, Data Studio, or Sales Copilot. If your subscription includes any one of these modules, you already have access, with no separate purchase required.

Why use AI Outbound scoring?

Most revenue teams prioritize outbound accounts manually via a spreadsheet list, a filter in some other tool, or intuition. That list goes stale fast, and when the person who built it leaves, the logic behind it leaves too. Ops ends up assigning accounts to sellers based on what's already sitting in the CRM, not on what's actually happening in the market.

AI Outbound Scoring changes that in a few concrete ways:

• You work from the full market, not just your CRM. Scores run against HG's constantly growing universe of 50 million+ accounts, so your team can find the best accounts to go after, including ones no one has ever touched.

• Sellers only see qualified accounts. Typically, only your top-tier accounts get assigned to reps, filtering out noise before it reaches the team.

• Sales and marketing work from the same list. Scores surface in the same place for both teams, instead of separate exports built at different times by different people.

• The model isn't a black box, and it doesn't leave when someone does. Every signal, weight, and scoring reason is documented and visible. Anyone on the team can open the model, read the logic, and adjust it.

• You can run more than one model. Different offerings or plays (e.g., a general ICP play vs. a competitor-displacement play) can each run their own model with their own signals and produce their own ranked account set.

• The list stays current. As a company's tech stack changes or it starts researching your category, the model picks that up automatically — it's not reflecting last quarter's research.

What makes the underlying signal different from other scoring approaches: it's built on HG's proprietary technology install, IT spend, and buyer intent (TrustRadius) data — not something available through CRM exports, generic AI tools, or firmographic-only providers.

What type of scores it returns

For every scored account, AI Outbound Scoring returns:

• A Fit, Need, and Intent breakdown, each backed by the specific conditions and point allocations that produced it, so you can see exactly why an account was ranked the way it was

• A composite tier, typically ranked A through F, that rolls the three dimensions into a single prioritization signal for sellers. Teams commonly route only their top tiers (A and B) to sellers or their marketing campaigns.

What data can be used in the scoring model?

Here is the list, you can ask the AI Agent in the screen at any time as well:

Fit

• Industry

• Number of Employee

• Revenue

• Industry

• Country, state, region

• Multinational presence (number of countries with operations)

• Fortune 500 / Forbes 2000

• Total IT budget

• IT spend as a % of revenue

• IT spend for a specific category

• Total AI budget

• AI budget as a % of IT budget

• AI spend for a specific category

Need

• Specific products installed (any technology in HG's catalog)

• Number of products installed from a defined set (stack fragmentation)

• Recent installs / install momentum (last 12 months)

Intent

• Research intent topics (what topics the company is actively researching online)

• TrustRadius Buyer activity (pricing views, comparisons, review page visits)

How to get started

You do not need Salesforce or any other integration connected to do any of this. AI Outbound Scoring runs on HG data and can the scores can be exported via the segment Export (coming soon: In Salesforce)

1. Go to Companies in RGI Platform

2. Click Add Scoring  

3. Click Create scoring profile

   1. 

4. Select “AI-assisted outbound scoring”

5. Enter your company domain and optionally give a specific goal the agent should aim for and/or a target account list to identify lookalikes from.
   

   1. The agent will research your company and build 3 different models targeting your different offerings

   2. Once created, select the ones you want to save

6. Review the first pass Look at the top accounts the model surfaces, sanity-check the Fit, Need, and Intent breakdown, and flag anything that looks off.

7. Click Publish live. it will take a few seconds

8. Click “Score Companies”. It will open the list of HG companies with the score applied to all companies

9. To filter the company list on Companies by Tier (A,B, C, D..) or Score (0-100), click on +Add Filter then under Scoring the filter Fit/Need/Intent
   

10. To view the details of a company score, click on a company score to see the score details
    

11. To adjust the scoring model, click on Change scoring

How to iterate on your scoring profile

AI Outbound Scoring is built to be adjusted, not set once and left alone.

• Update the model in seconds. As your ICP, competitive landscape, or market shifts, you can change signals and weights easily interacting with the AI agent in the chat. You can

  • change disqualifiers

  • update points

  • add / drop conditions  (e.g. an industry value, a product_

  • add dimensions (e.g geography, spend ….)

  • change tiers buclet

  • change the Fit , Need, Intent weights

  • rename dimensions and sub dimensions

• Correct the model when something looks off. If the AI's research on your company or a signal doesn't look right, just ask the AI agent to correct it and the model updates accordingly, creating a draft

• Run separate models for separate plays. Rather than trying to make one model do everything, configure a distinct model per product, offering, or motion (e.g., new-logo ICP targeting vs. competitor displacement), each with its own signals and its own ranked output.

Where to use the scores

• Companies: As a filter to create Segments and export companies
  

• Platform Export: export your segment with the scores

• Salesforce: available with HG Force package version 1.18 and above. Follow this article to install

• Hubspot: coming soon

• AI Sales Copilot: scores of live models automatically show to your sellers on their book of accounts. Learn more here.

• API: search and score companies programmatically via API using the scoring models you configured in the Platform. Follow this article for more details

• MCP: coming soon

• S3 / Databricks / Snowflake: coming soon