A sales rep spends 20-30 minutes researching a single prospect. Company background, recent funding, tech stack, key contacts, competitive landscape, relevant pain points. Multiply that by 50 new leads a week and you’ve burned a full-time role on work that never touches a phone call.
The instinct is to skip the research. Send generic outreach. Hope the volume makes up for the lack of personalization. It doesn’t — response rates drop, and the leads that do respond get a rep who’s underprepared.
There’s a middle ground: automate the research, keep the human on the relationship.
What the New Lead Pipeline actually does
The Sales starter pack includes a workflow called New Lead Pipeline. It takes a lead’s name and company, then runs three steps:
-
Prospect Research — The AI searches for company information, recent news, funding history, leadership team, and tech stack. The output is structured: company overview, key contacts, recent events, and competitive positioning. Not a wall of text — labeled fields your rep can scan in 30 seconds.
-
Lead Qualification — The research output feeds into a scoring recipe that evaluates the lead against your ICP criteria. Company size, industry fit, tech stack alignment, budget signals. The result is a qualification score with specific reasons for the rating.
-
Condition: If Qualified — A condition step checks the qualification score. If the lead meets your threshold, the workflow continues. If not, it stops — no point drafting outreach for a poor-fit lead.
-
Draft Outreach — For qualified leads, the AI generates a personalized cold email and LinkedIn connection request using the research data. The email references specific company details, recent events, and relevant pain points. It’s not a template with a name swapped in — it’s a draft built from actual research.
The entire pipeline runs in under a minute. A rep reviews the output, makes any adjustments, and sends.
What the human still controls
The workflow generates drafts. It doesn’t send emails. It doesn’t add contacts to your CRM. It doesn’t make qualification decisions for you — it presents a score with reasoning that a rep can agree with or override.
This matters because:
- Qualification criteria evolve. Last quarter you targeted mid-market SaaS. This quarter you’re moving upmarket. The rep adjusts the threshold, not the AI.
- Outreach tone is personal. The draft gets you 80% there. The last 20% — the aside about a shared connection, the reference to a podcast the prospect was on — that’s the rep.
- Deals are relationships. Automating research scales your capacity. Automating the relationship kills your pipeline.
Running it on a schedule
Most sales teams don’t run the pipeline one lead at a time. They set up a schedule that pulls new leads from a Google Sheet or CRM export every weekday morning at 8 AM. By the time the team opens their laptops, every new lead has research, a qualification score, and a draft email waiting.
The schedule supports multiple input modes. You can pull from Google Sheets directly, hit a webhook to pull from your CRM’s API, or use static values for testing.
Customizing the recipes
Every recipe in the pipeline is fully editable. The Prospect Research recipe has a prompt template that you can adjust to focus on what matters to your team. If you sell to healthcare companies, you might add instructions to look for HIPAA compliance signals and regulatory news. If you sell developer tools, you might emphasize tech stack and engineering blog posts.
The output schema is also customizable. If your team needs a field for “estimated deal size” or “buying committee members,” add it to the schema and update the prompt.
Beyond prospect research
The Sales pack includes workflows for the full sales cycle:
- Discovery to Proposal — From call notes to a polished proposal executive summary
- Deal Review Pipeline — Weekly pipeline review with risk flagging and action items
- Competitive Positioning — Analyze competitors and generate positioning strategies
- Account Renewal — Health check, renewal brief, and proactive outreach for existing accounts
Each one follows the same pattern: automate the research and drafting, keep the rep in control of the decisions and relationships.