Account plan for <company.com>

The problem this solves

The discovery call is tomorrow and the account record holds a domain, an industry and a note from a conference two years ago. A real plan from there means a dozen tabs: the company site, a database for headcount and offices, a technology lookup, the careers page, a funding tracker, a news search, a people search and an email tool. Reps do the full version for the two accounts they care most about and skip it for the rest, so plan quality tracks the calendar instead of the account's value.

A plan is only useful when every claim in it can be checked. A priorities section that says the company is investing in data has to point at the postings and the round that suggest it, and a buying committee has to name the people who hold those roles today, with addresses that will deliver. A plan written from memory reads well in the doc and falls apart in the first five minutes of the call.

How the mission runs

1. Resolve the domain and pull the firmographics. The Company Intelligence Agent resolves <company.com> to one company record and pulls firmographics: what the company sells in its own words, headcount and its trend, HQ and other offices, and the industry. Each figure carries its source, so the doc can show where a number came from.
2. Read the stack with technographics. Technographic data lists the technologies detected on the company's properties, grouped by function, with the date each was first seen. The agent keeps the categories relevant to your product and notes anything that overlaps with or competes against what you sell.
3. Add hiring, funding and news as timing evidence. Hiring data supplies the open roles and how the count has moved; funding and news data supplies the rounds, their investors and the latest announcements. Read together they become the likely priorities: a team hiring three data engineers after a round has a project underway, and the plan says so with the postings and the round linked.
4. Assemble the buying committee and verify the emails. People search finds the people at <company.com> who hold the committee roles you describe - economic buyer, likely champion, technical evaluator. The email finder runs as a waterfall, one source after another until an address is found, and email verification checks each address before it enters the plan, so every person shows verified, risky or unknown.
5. Write the angle from the evidence. With the stack, the hiring pattern and the news side by side, the agent drafts the angle: which apparent priority your product serves, what in the stack it complements or replaces, and the two or three facts a first message should reference, each linked to the evidence it rests on.
6. Deliver the Google Doc and the #sales summary. The plan is written to a Google Doc with sections in the order of your prompt, and a five-to-eight-line summary is posted to #sales with a link. Nothing goes to the account itself; the plan is a brief for your team, and the run is logged with the credits it used.

The prompt

This is the exact objective the agent receives. Swap the obvious placeholders for your own domain, segment or channel and run it as-is from the console, Slack, or the API.

What comes back

A Google Doc named for the company: what <company.com> sells, headcount and locations, the stack with first-seen dates, open roles and the hiring trend, funding rounds and investors, recent news with links, a buying committee table with titles, emails and verification status, the likely priorities and your angle, each claim with its source. The #sales post carries the headline and links to the doc.

Make it yours

• Name the committee explicitly - VP Engineering, Head of Data, CFO - and the agent looks for those titles instead of inferring them.
• Ask for phone numbers alongside emails when this is an account your team calls; the phone finder runs the same waterfall.
• Point the summary at a deal channel or a Notion page instead of #sales, or drop it and keep the plan in the doc.
• Run it across the ten accounts on list <name> to get ten docs and one #sales thread with a message per account.

Frequently asked questions

Go deeper

• Company and person data for AI agents: the complete guide
• The complete guide to agentic AI for GTM data
• Outbound personalization that references evidence
• Technographic and hiring-signal prospecting: the complete guide