Watch every deal. Act on the unhealthy ones.

What is AI deal health monitoring?

AI deal health monitoring is the continuous evaluation of open sales opportunities against signals that predict stalling, slippage, or revenue risk. Instead of showing those risks in another dashboard, Ollie turns them into prepared work.

The work that moves a deal forward sits in the gap between meetings. Top performers consistently send the re-engagement email, answer the procurement question, and loop in the new stakeholder. But across a full pipeline, every team has a long tail of deals where attention has moved on and the next step is waiting for someone to remember it.

The Supermetrics revenue team runs Ollie across 50 sellers today. Read the full case study.

What is deal scoring?

Deal scoring rates every open opportunity on whether it is moving forward or at risk, so a team knows which deals need attention first. A score built only from CRM fields knows what reps remembered to enter. The deal health score behind Ollie’s alerts is built from the deal itself: every email, meeting, and CRM update on the opportunity. It is a diagnosis of where the deal stands, not a prediction of whether it will close.

The score combines 6 weighted criteria: engagement, deal movement, fit, risk, financial signals, and recent interaction momentum. What the buyer says and does counts for more than what the rep records, and recent signals count for more than older ones. Each criterion shows the evidence behind it, so nobody has to take the number on trust.

When a deal’s health drops, Ollie tells the deal owner in Slack or Teams which criteria moved and recommends what would bring the deal back.

What does an AI sales co-worker check on every deal?

Ollie watches for 5 common sources of pipeline risk: a missing next step, stale activity, a qualification gap, a health-score drop, and a close date at risk.

1. Next steps. The deal needs a clear next action.
2. Stale deal re-engagement. A deal without meaningful activity.
3. MEDDPICC or SPICED gap. A qualification field empty in a stage where it should be locked.
4. Health score drop alert. A deal with a drop in health.
5. Close date approaching. The close date is near and blockers remain.

Check 1: next steps

Ollie finds the next action that keeps the deal moving between meetings. Ollie uses deal context, past conversations, open follow-ups, and agreed next steps to identify what should happen next.

That next step could be sending the promised follow-up, booking the next meeting, asking the Paper Process question, looping in a stakeholder, or updating the deal after a buyer reply. Whichever it is, Ollie drafts it and posts it to the rep in Slack or Teams.

Your buyer asked for line-item pricing on 3 integration options, and the next review is on Friday. The draft gives the per-line pricing for each integration path and asks which one to bring to that review.

Check 2: stale deal re-engagement

Ollie finds deals that have gone cold: no meaningful activity for long enough that the opportunity needs a new reason to restart the conversation.

Meaningful activity means emails sent or received, calls or meetings logged, deal updates, or buyer-initiated contact. Internal notes and passive system updates do not count. A deal can still be stale even if someone updated a CRM field.

For each stale deal, Ollie prepares a re-engagement angle from the deal context and attaches a draft email for the rep to review.

Their VP Engineering has been speaking publicly about platform velocity. The draft connects the talk to the pressure your contact described, offers a 1-page view of how other teams ran the same play, and asks for a 20-minute call.

Check 3: MEDDPICC or SPICED gap

Ollie finds qualification fields that are still empty in stages where they should already be locked. Common examples include Paper Process missing in Negotiation, or a deal entering Proof of Value before the rep has confirmed who owns the final buying decision.

Ollie is stage-aware. Instead of just flagging missing data, Ollie drafts the question that should be asked next, using the right stakeholder and deal context.

The draft asks their CFO who reviews the MSA, whether legal redlines are central or run by the business unit, and the rough timeline from agreed terms to countersignature.

Check 4: health score drop alert

Ollie flags a deal-health drop the day the score crosses the line, then stays silent until something changes.

The score combines 6 weighted criteria covering engagement, deal movement, fit, risk, financial signals, and recent interaction momentum. Customer-sourced data weighs more than internal data. What the buyer does counts for more than what the rep types into a field.

When the alert lands, the rep sees which criteria drove the drop and where the recovery work sits.

Behind each reason sits the evidence: the last inbound from their CFO was May 17, their CFO has been the slow lane on every recent contract, your champion is traveling and unresponsive, and the stage has not moved.

Check 5: close date approaching

Ollie catches deals that are nearing their close date with blockers still unresolved. Ollie checks the deal record, recent conversations, and open next steps, then lays out what could stop the deal from closing on time.

The deal owner gets the blockers, the context behind them, and a suggested next move for each. The rep can work through the list or push the close date with eyes open instead of letting it slip silently.

Each blocker comes with its context and a suggested move:

1. Legal redlines outstanding. The MSA went to their legal team on May 18, with no reply. Suggested move: a drafted nudge to their general counsel, to review and send.
2. Security review incomplete. The SOC 2 questionnaire came back on May 20 with 2 items flagged. Suggested move: loop in your security lead to clear both today.
3. Their CFO hasn’t countersigned. The CFO is in board prep this week. Suggested move: a drafted 1-line check-in to the CFO’s assistant to lock 15 minutes on Friday.

What managers see

Manager morning briefing. When a deal amount moves between briefings, Ollie flags it the morning after the change. Forecast risk reaches the manager the day after, not at the end of the quarter when the gap is too big to close. That is the start of forecasting with Ollie.

Friday CRM sweep. Ask Ollie to check deals touched that week and prepare suggested CRM updates for review. Reps and managers can set this up themselves; admins can roll it out across the team.

Win/loss analysis. Ollie reads every closed deal: emails, calls, meeting recordings, and CRM activity. Outcomes are classified across 13 reason codes and the MEDDPICC patterns behind them, and each pattern becomes a coaching point for the next 1:1. When a pattern repeats across open deals, it is a pipeline bottleneck.

Together, they turn the weekly pipeline review into a conversation about the deals that need it.

Why does this work as a co-worker and not as another dashboard?

Most deal-health dashboards assume reps will check them. They don’t. The information was already in the CRM. The problem was never visibility; the problem was that nobody acted on it.

Ollie works the other way around, noticing the risk, preparing the work, and bringing it to the rep. That is what makes Ollie an AI sales co-worker, not an assistant. An assistant waits for the rep to know what to ask. A dashboard waits for the rep to remember to check. A co-worker notices first, prepares the next step, and brings it to the person who can decide.

Where Ollie fits in your stack

Ollie connects to 3 parts of your sales stack. The checks run on Optivian’s deal intelligence: a current understanding of each deal, built from every email, call, meeting, and the CRM record as they land.

Conversation intelligence, the source

Ollie captures meetings natively. If a recorder like Gong or Fireflies is already in place, Ollie reads what those tools have logged to the CRM instead. Either way, Ollie has the meeting context the work needs.

CRM, the system of record

Optivian reads the deal record and its activity (notes and call summaries) from Salesforce, HubSpot, or Pipedrive, including whatever the conversation intelligence tool has logged there, and accepted updates write back to the same record.

Generalist AI, the surface

Reps already ask Claude and ChatGPT for help on deals. Through the connector Optivian ships, those tools can read the deal context Ollie keeps, so the answers come back grounded in what is actually happening on the account.

Ollie is the co-worker who turns deal context into action.

What does this look like running at a 50-seller revenue team?

Supermetrics: AI deal health monitoring across 50 sellers

At Supermetrics, Ollie was deployed across 50 sellers and ingested thousands of customer interactions: emails, calls, meetings, and CRM updates. The team saw strong adoption and measurable pipeline activity within 30 days.

96% weekly active users. 48 of 50 sellers opened, reviewed, or acted on something Ollie posted. Adoption did not depend on a rollout campaign or mandatory training; reps started using Ollie because the work showed up where they were already working.

72 stale deals reactivated. Each had sat untouched for at least 2 weeks before Ollie restarted the conversation.

Read the full Supermetrics case study.

Related: the same deal picture drives the prep before every deal meeting, and the follow-up after it, and the mutual action plan your champion carries in. For the manager’s side of the same checks, explore pipeline reviews and forecasting.