One domain in. A scored, cited account list with decision makers out.
Give your AI an ICP. It researches accounts in your browser, scores each with evidence you can click, and names the buying committee.
“Score these accounts against my ICP of mid-market B2B SaaS with a RevOps buyer, with evidence, and name the buying committee.”
Northwind
northwind.io
Halcyon Labs
halcyon.dev
Tidemark
tidemark.co
From your ICP to a scored account brief
Connect the AI you already use, whether that’s ChatGPT, Claude, or Gemini, to your browser. From your ICP it runs four steps and stops before anything writes to your account plan.
- 1
Hand it your ICP, not a query
Describe the accounts you want by segment, size, motion, and a few of your best customers. Your AI turns that into a working profile it can score against, no filter syntax to learn.
- 2
It researches each account in your browser
It opens the real surfaces a researcher would: the company site, pricing and careers pages, recent news, and the LinkedIn org page. All of it runs in your own logged-in Chrome, not a cached firmographic feed.
- 3
It scores accounts on live signals
Every account gets a fit score with the evidence attached: “hiring 4 SDRs,” “opened an EU office,” “shipped usage-based pricing,” each pulled from the page it actually read.
Each score links back to the page behind the signal.
- 4
It names the committee and writes the brief
For top accounts it identifies the buying committee by role and assembles an account brief you can take into a pipeline review. Writing it to your account plan waits for your approval.
Your AI wants to add 5 scored accounts and 9 contacts to your account plan sheet.
+ 2 more accounts · 9 contacts staged for export
It assembles the plan. The export stays yours.
Writing scored accounts and named contacts into your account plan is a real action on your system of record, so it works like every consequential Actionbook action: you see exactly what the agent wants to add before it commits, and nothing lands without your approval.
- Queued, not auto-written: The default stopping point is a reviewed export, with accounts, scores, and contacts staged for your okay.
- Approve the write: The full payload surfaces for approval first, in plain language, before it touches your sheet or CRM.
- Watch it work: Everything happens in your own Chrome, so you can inspect a source page or take over at any point.
Built for account-based teams
One workflow, a scored and cited account list, tuned to how each team plans coverage.
ABM marketers
Build a tight, evidence-scored account list for a play, then hand sales briefs that already name the committee and the signal behind each pick.
Enterprise AEs
Plan a territory by scoring named accounts on live hiring, expansion, and product signals instead of a year-old firmographic snapshot.
RevOps teams
Tier accounts with a transparent, repeatable fit score where every field links to its source, so the model is auditable instead of a black box.
Evidence you can open
A score you can’t trace is just a guess. It attaches the live page behind every signal, the careers post or the pricing change, so each number traces back to a page you click, not a static intent feed.
Founders
Pick design partners by researching a short list of accounts end to end, covering site, pricing, hiring, and news before you spend a single sales hour.
Prompts to start scoring accounts
Copy a prompt, paste it into your AI, and Actionbook runs it in your own browser session.
Score these 15 accounts against my ICP, which is mid-market B2B SaaS with a RevOps buyer. For each, give a 0–100 fit score with one line of evidence per score and a link to the page you read it from.
Build the buying committee for Northwind from its company site and LinkedIn org page: list the likely economic buyer, champion, and blocker by role, with what each one would care about.
Sweep the careers pages of these eight domains for expansion signals like new sales, RevOps, or data roles, and flag any account hiring three or more on a team that matters to my play.
Draft a one-page account brief for my Monday pipeline review on my top three accounts: fit score, the signals behind it, the committee by role, and a recommended angle for each.
Open the pricing pages for these accounts and tell me which moved to usage-based or seat-based pricing recently, since that changes how we’d position. Cite the page for each.
Re-check my Tier 1 account list for new funding, leadership changes, or fresh job postings since last month, and give me just the accounts where something moved, with the source.
Frequently asked questions about GTM intelligence
What account-based teams ask before pointing an AI at their ICP and account plan.