Clay Automation: Pre-Call Research & Talking Points Generation

Overview

Clay offers a built-in template called "Automated Pre-Call Research for AEs" that uses AI to generate personalized talking points and company summaries before sales calls. This article covers how to set up, customize, and integrate this template into the Asymmetric sales workflow.

The core value: instead of reps manually researching each prospect, Clay scrapes company websites and other sources to produce ready-to-use call prep notes — reducing prep time and improving call quality.

The Template: Automated Pre-Call Research for AEs

What It Does

The template creates a structured Clay table with the following enrichment columns:

Column	Source / Method
Name, Email	Input data
Identify Email Type	Clay enrichment
Domain	Derived from email
Person Enrichment	Multi-source waterfall
Title	Enriched
Company Enrichment	Owler, ZoomInfo, etc.
Founded Date, Description	Company data
Technology Stack	BuiltWith
Website Summary	AI prompt against company website
Talking Points	AI prompt (customizable)
Customer Names	AI scrape from company website

How to Access It

In Clay, search for the "Automated Pre-Call Research for AEs" template.
Click "Duplicate this template" to save a copy to your workspace.
The duplicated table will appear in your Clay workspace, ready to customize.

Note: Clicking the template link may automatically duplicate it to your workspace. Check your workspace before duplicating again.

Key Columns: Website Summary & Talking Points

Website Summary

Type: AI-generated
Prompt pattern: "I'm a sales rep selling to [account]. Give me a summary of this website."
Input: Company website URL (pulled from enrichment)
Output: A concise overview of what the company does — useful context before the Talking Points column.

Talking Points

Type: AI-generated (fully customizable prompt)
Default prompt pattern: "I'm a sales rep selling [accounts] available automation software. Give me talking points for a call."
Output: 3–5 targeted conversation starters based on the company's public presence.

Customizing the Talking Points Prompt

The prompt can be edited to surface specific intelligence. Examples:

Find customer names: "Scrape their website and find any customer or client names they mention — e.g., 'Some of our best customers include...'"
Find news/thought leadership: "Find any recent blog posts, interviews, or articles published by [first name] [last name] at [company]."
Competitive intelligence: "Identify any competitor mentions, funding rounds, or hiring trends visible on their website or in recent news."

To edit: click the column header → Edit Column → modify the AI prompt text.

Recommended Workflow

Clay Table (contacts + company URLs)
        ↓
Run enrichment columns (person, company, tech stack)
        ↓
Website Summary column populates
        ↓
Talking Points column populates
        ↓
Export / sync to HubSpot contact record
        ↓
Rep reviews Talking Points before dialing

Integration with HubSpot

Talking Points and Website Summary outputs can be mapped to custom HubSpot contact properties.
This allows reps to see pre-call research directly in HubSpot without switching tools.
See ^[1] for property configuration details.

Columns to Add or Remove

When customizing the duplicated template, consider:

Remove (if not relevant):
- Owler enrichment (if not subscribed)
- BuiltWith technology stack (if tech stack isn't a qualifying signal)

Add:
- Customer Names — AI scrape to identify named clients on the prospect's website; useful for social proof in calls ("I see you work with X — we also work with companies like them")
- Recent News — pull recent press mentions or announcements
- Headcount — useful for sizing the opportunity

Data Requirements

The template works best when input records include:

Field	Required?	Notes
Company Website URL	Strongly recommended	Drives Website Summary and Talking Points
Email	Recommended	Enables person enrichment waterfall
First / Last Name	Required	Used in person-level prompts
Company Name	Required	Used in company enrichment

For lists with missing emails (e.g., the initial 159-contact GL Food & Beverage list), run ^[2] first to fill gaps before running the pre-call research template.

Practical Notes

Prompt quality matters: Vague prompts produce generic output. Be specific about what the rep needs to know before a call.
Credits: Each AI column run consumes Clay credits. Prioritize columns that deliver the most call value (Talking Points > Website Summary > others).
Iteration: The template is a starting point. Edit prompts after reviewing initial outputs — if talking points are too generic, tighten the prompt with more context about your ICP or value proposition.
Asymmetric value prop context: Consider injecting the asymmetric value proposition framing into the Talking Points prompt so AI-generated points naturally align with the sales narrative. See ^[3].

^[2]
^[1]
^[3]
^[4]
^[5]