The problem this solves
One sheet, eleven rows: <company.com> at the top and the ten companies most like it underneath, with the same four measures for every row. Headcount and its trend, total raised and the last round, a stack summary, and how many roles are open now against ninety days ago. Read across a row and you know a company; read down a column and you know the field.
The reason this rarely exists is that the four measures live in four systems and the set itself has to be built first. A rep asked who a prospect competes with lists the names from the last call; a founder asked the same question lists the names from the last board meeting. Neither list contains the company that was founded two years ago and is now hiring faster than anyone else in the set.
Side by side is the format that makes the numbers mean something. Headcount alone says little; headcount against total raised says who is efficient, who is ahead of their funding, and who is about to need more of it.
How the mission runs
- Build the set with similar companies. The Company Intelligence Agent resolves <company.com> and asks similar companies for the ten nearest matches, weighting what the company sells, the customers it names, headcount band and geography. Each candidate arrives with a reason, and the set is listed at the top of the sheet.
- Fill the firmographics. Firmographics give every row its headcount with a twelve-month trend, HQ, founding year and industry. Missing fields go through waterfall enrichment, so a source is only charged when it fills the gap.
- Add funding. Funding and news data supplies total raised, the last round with its date and stage, and the lead investor. Rows with no disclosed funding are marked as such rather than shown as zero.
- Summarize each stack. Technographics run against every domain in the set. Instead of forty detections per row, the sheet shows the CRM, marketing platform, analytics and infrastructure choices, with a count of detected technologies and a note when two companies in the set share the same core tools.
- Measure hiring momentum. Hiring data returns the open roles at each company now and ninety days ago. Momentum is the change between them, and the sheet also lists the functions where the openings cluster, since a company adding sales roles and a company adding engineers are in different phases.
- Chart and deliver the sheet. The chart capability renders a scatter of headcount against total raised with one point per company and <company.com> highlighted. The Google Sheet holds the comparison on one tab, the evidence links on another and the chart on a third.
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 Sheet with eleven rows - <company.com> and its ten closest companies - and columns for headcount and trend, total raised and last round, a stack summary, open roles now and ninety days ago with the momentum figure and the hiring functions, plus a scatter chart of headcount against total raised with the subject company marked. Every cell that came from a lookup links to its source.
Make it yours
- Ask for twenty companies and a fit score column, so the set doubles as a target list for the Prospecting Agent.
- Replace hiring momentum with intent data when you care more about which companies are researching your category than which are hiring.
- Save the set as a refreshable list and rerun the comparison monthly to see who moved.
- Deliver it to Notion as a database instead of a sheet when the team keeps market notes there.
Frequently asked questions
Why did the agent pick these ten?
Similarity is measured on what a company sells and to whom, plus size and location, and the reasons are listed per row. Add or remove names in the prompt and the comparison runs on your set.
How is hiring momentum calculated?
As the difference between open roles now and open roles ninety days ago, from hiring data, with the functions where the change happened listed beside it. Change the window in the prompt if ninety days is wrong for your market.
Can it run again next quarter?
Yes. Save the set as a list and rerun the prompt; the sheet is rebuilt with a column showing the change in each measure since the last run.