How AiSDR Personalizes Outreach: Data Sources Explained

This article explains what data AiSDR uses to personalize messages, how each source works, and how to make sure the right data is available before you launch a campaign.

Data sources at a glance

AiSDR can pull personalization signals from five places:

• Live AI research: real-time web search tied to the signals you set up for lead search in Live AI database (facts that are less than 6 months old)

• Web research: any recent company or lead news available in web

• LinkedIn: profile info and recent posts

• CRM data: past email engagement, meeting notes, and deal info from HubSpot or Salesforce if connected

• Social signals: who's reacting to or commenting on relevant LinkedIn posts

How it all comes together

Once data is collected and stored on the lead's profile, the system hands it to the AI as a single research dossier with one instruction: personalize the message based on the most relevant piece of information available.

Each step in a campaign sequence can have personalization toggled on or off, and different campaigns can lean on different data sources depending on what's available.

The AI picks the strongest hook and anchors the message there - the same way a good salesperson would skim a brief and lead with the line that lands hardest.

What tends to surface

• HubSpot notes win when there's real interaction history (past meetings, email replies, deal context) - hard to beat for relevance.

• Web research wins when there's a fresh, timely signal that matches your offering (funding round, product launch, recent press).

• LinkedIn data is the fallback for role- and company-based personalization when nothing more specific is on file.

When Live AI is involved: Live AI signals are always prioritized over the other sources.

How each personalization source works

1. Live AI

When you set up a Live AI search with specific signals (like “companies hiring for AI roles”), the system stores the exact criteria that matched, the reasoning for why this lead qualified, and a relevant snippet from the source. It also generates a company summary.

All of this gets wrapped in dedicated tags and the AI is explicitly told to prioritize this info for personalization. So unlike other sources where the AI picks what's relevant, Live AI leads come pre-loaded with the reason you wanted to reach them - making the personalization much more targeted and signal-specific.

If you're sourcing leads through Live AI, set up clear signals first - those signals are the personalization anchor the AI will lean on. 📄 How to create a good prompt with Live AI feature?

2. Web Research

The system does a live internet search on the lead using their email, LinkedIn URL, and company info. It finds up to 3 recent, verifiable facts (less than 6 months old) and explains why each is relevant. Results are cached for a month so we don't re-search the same person.

3. LinkedIn Profiles

We pull the lead's profile info (headline, summary, job history) and their recent posts. We have multiple providers for this, and the system falls through them in priority order based on what's enabled. If one provider's data is stale or unavailable, it tries the next.

4. HubSpot Fields & Notes

If your Hubspot is connected to AiSDR, we can pull different info categories from the CRM: basic lead info (company, title), email engagement history, notes/activity logs, and meeting history. All of this gets bundled and passed to the AI as context.

Learn more on how to do that here: 📄 HubSpot + AiSDR integration

5. CSV Personalization Column

When uploading leads via CSV, there's a free-text “Personalization info” column option you can upload. Whatever you write there (e.g., “Downloaded our pricing guide, visited demo page”) goes directly to the AI as context. It's completely manual - you decide what goes in. 📄 How Can I Add Additional Information for AI-Powered Email Personalization?