ABM Hub - Deal health (components, evidence and AI recalculation)
Deal health measures how solid an opportunity is according to the components of your product's health framework (for example identified pain, decision criteria, decision process, competition, economic buyer). It is evaluated per opportunity, not per account. In the Target Account score, the product'
# ABM Hub - Deal health
📘 What it is
Deal health measures how solid an opportunity is according to the components of your product's health framework (for example identified pain, decision criteria, decision process, competition, economic buyer). It is evaluated per opportunity, not per account. In the Target Account score, the product's open opportunity on that record drives it (if there are several, the one with the highest amount).
Each component has a level:
- 0 Unknown
- 1 Hypothesis
- 2 Hint
- 3 Confirmed
- 4 Confirmed and documented
The AI proposes up to "Hint"; only a person confirms. Levels 3 and 4 require evidence: who said it and where it is recorded; level 4 also asks for a document or a second source. Your record takes priority: the AI can no longer change a component that a person recorded.
Indicators at the top: Health framework, Score (weighted average of the components that apply to the phase, from 0 to 100), Confidence (percentage of applicable components at 3 or 4 that the prospect confirmed) and Last calculation.
Two blocks: Apply in prospecting (committee evidence, stored on the account and shown the same in all its open opportunities for the product) and Apply from "With opportunity" (each opportunity's own evidence; without an open opportunity they are not evaluated).
📍 Where to find it
Intelligence > ABM Hub > Target Accounts > open the account > Deal health tab. There is also a health badge on the opportunity record.
🛠️ How to do it
Evaluate a component (without evidence, up to Hint)
- In the component's row click Evaluate.
- Choose the Level, write What we know and, if you want, Who said it. Click Save.
Confirm a component with evidence
- Click Confirm.
- Choose New level (3 or 4), Who said it (a contact of the account) and the Evidence: From an interaction (only this account's appear), Paste the quote (exact quote, where and when) or Attach a document (PDF, DOCX, XLSX, PNG or JPG up to 10 MB; an email is uploaded exported as PDF or image).
- For level 4 also add a Second source.
- Click Confirm.
View the history
History shows each change, who made it and whether it was the AI.
Recalculate health with AI
- Click Recalculate. The AI reads the contacts' conversations, notes, calls, meetings and transcripts, the account's research and learning and, if you allow it, the attached documents.
- Choose Opportunities (only the selected one or all the product's open ones) and What to read (since the opportunity was created, since the first enrollment in its origin campaign, or the whole history). You can check Include attached documents (it costs per new document).
- Review Total cost, Available and Would remain and click Recalculate. You are charged per opportunity and only if the AI finishes successfully. The AI does not change what a person already recorded.
- You can also do it for several accounts from the listing with Recalculate health (all or nothing depending on your balance).
The components that apply in prospecting are also filled in automatically, at no cost, from confirmed learning facts and map roles (never beyond "Hint" and never overwriting what a person recorded).
❓ Frequently asked questions
Why does it say "Needs an open opportunity" / "Without an open opportunity, health stays at "—" in the account score"? Because health is evaluated per opportunity. Create or link an open opportunity for the product.
Who can evaluate and confirm? Whoever works the account; with other roles you will see "Read only".
Why can't I go past level 2? Without evidence you cannot go past 2: you need who said it and where it is recorded.
What are "Archived evaluations"? Evaluations that no longer count toward the score; they are never deleted.
Am I charged if the AI finds no hints? Yes, if the AI ran and finished successfully, even if it proposes no new hints (the notice shows the credits charged). If it fails, you are not charged.
🔧 If something doesn't work
- The Evaluate or Confirm button is disabled. An open opportunity is missing for the opportunity components, or you are not whoever works the account.
- "Your balance covers N opportunities. All or nothing". Reduce the scope or top up credits.
- The score didn't move after recalculating. The AI only proposes hints (level ≤ 2); the score rises when a person confirms with evidence.
- Escalate to a person with: company, user, account, opportunity, URL, time, steps and a screenshot.