Deal risk is the chance that an open opportunity will not close on the terms and date in the forecast. In practice it almost always comes down to one thing: a piece of buyer evidence the deal never collected, such as a signer who has never heard the case or a date the buyer never agreed to.
That definition changes what you do about risk. If risk is a score, you watch it. If risk is a missing piece of evidence, you go and get it, in a conversation with a named buyer, this week.
How this guide is organized
| Section | The question it answers | Go deeper |
|---|---|---|
| What deal risk is | What are we measuring? | What is a deal risk score? |
| Where risk comes from | Which evidence is missing? | How to coach a stalled deal |
| How tools detect risk | What do Gong, Clari and others flag? | Deal risk score vs deal coaching, Clari vs Gong |
| Pipeline coverage | Is there enough real pipeline? | What is pipeline coverage? |
| Close dates | Why do dates slip, and what does the manager say? | How to coach reps on close dates |
| Qualifying out | When should a deal leave the book? | How to qualify out a deal |
| The forecast call | How is risk priced into the number? | How to run a forecast call, Forecast call agenda template, How to improve forecast accuracy |
| From flag to fix | What does the manager do with an at-risk deal? | Deal coaching guide |
What deal risk is
Deal risk is the gap between where a deal sits in the CRM and what the buyer has done to put it there. A deal in "Proposal" with no economic buyer on any call is at risk, whatever the score says.
Most definitions describe risk as warning signs: the champion goes quiet, the thread narrows to one contact, the close date moves. Those are real, and they are symptoms. A champion goes quiet because the deal lacks something the champion needs to move it internally. A close date moves because the date was the rep's and the buyer never tied anything of theirs to it.
So this guide uses a working definition with two parts:
- The evidence test. For each stage, there is a piece of buyer evidence that should exist: something a named buyer said, did or agreed to. A deal is at risk when that evidence is missing at the stage it claims.
- The time test. A deal is also at risk when the buyer's last action is old and no dated buyer step exists in the future, even if the evidence was collected once.
Both tests take a few minutes per deal, with or without software.
Where deal risk comes from: six missing pieces of evidence
Nearly every at-risk deal is missing one of six pieces of buyer evidence. The table maps the signal a tool or a manager sees to the evidence underneath it, the question that tests it, and the conversation that fills the gap.
| Risk | The signal you see | The missing evidence | The test question | The conversation that fixes it |
|---|---|---|---|---|
| 1. No economic buyer | Champion says "it is with leadership"; nobody senior on any call | The person who signs has never heard the case or asked about price or timing | "When was the signer last on a call, and what did they ask?" | Champion call: build the two points the signer will ask about, and get a date for the signer to hear them |
| 2. No cost of the problem | Buyer calls the product "interesting"; meetings slip | No named buyer has said what the problem costs them | "What is the number, and who said it?" | Ask the person who described the problem to put a rough figure on it with the rep |
| 3. Single-threaded | One contact on every call and email | Nobody else at the account has a reason to want the deal | "Who else loses if this does not happen, and have they told us?" | Ask the champion who else feels the problem, and for an introduction framed around their part of it |
| 4. Unmapped buying process | "It is with procurement" for weeks, no named person | The rep cannot name the approval steps in order, with owners | "Name every step and person between now and a signature." | Ask what happens the day after a signed contract arrives, and who it goes to first |
| 5. An unaddressed alternative | Late feature questions, a request for a comparison matrix | Another vendor, an internal build or doing nothing is in the deal and nobody has named it | "Has another option come up in the last three conversations?" | Name it with the buyer and ask what the other option would have to be better at |
| 6. A rep's date | Close date moved twice; buyer friendly and unhurried | The buyer never named a date their own plan depends on | "What is on their calendar, and who put it there?" | Ask for the buyer's date and the event behind it |
Two rules make the table useful. First, run it in order: the earliest missing piece is usually the one that broke the deal, because everything later depends on it. Second, a rep rules a row out only with buyer evidence, meaning a quote, an email or a meeting the buyer accepted. "I am sure the CFO is on board" rules nothing out.
The same six pieces sit inside every major qualification method, under different names. MEDDPICC calls the first one the Economic Buyer and the fourth the Decision Process and Paper Process (MEDDPICC). Gap Selling puts the cost of the problem at the center of the deal (Gap Selling). Whatever method your team runs, the risk table is that method read backwards: start from what is missing and work out the conversation.
How tools detect deal risk
Revenue intelligence and forecasting tools detect deal risk by reading activity and conversation data and flagging deals whose patterns look like past losses. They are good at the signal column of the table above and say little about the conversation column.
Gong describes its forecasting product as one that "detects deal risk in real time, so you can spend more time steering pipeline forward". Clari Forecast says it lets a team "act early on slipping deals, stalled buying activity, and execution gaps" and that risks "automatically trigger next best actions in the seller workflow". Avoma bundles its version as Deal Health, tracking signals such as single threading, account activity and close-date changes.
The output is a flag on a deal, often a deal risk score. That is useful for triage: it tells a manager with 80 open deals which ten to look at first. It cannot tell the rep which buyer to call and what to say, because the score is built from the signal column and the fix lives in the evidence column. We covered why reps learn to scroll past red badges in a risk score is not a fix.
Pipeline coverage hides risk inside a ratio
Pipeline coverage is the value of open pipeline divided by the target for the period. It tells you whether there is enough pipeline on paper, and it treats every dollar in the CRM as equally real.
A team at a comfortable-looking ratio can be carrying half its pipeline in deals that fail the evidence test. The fix is to calculate coverage twice: once on everything, and once only on deals that pass the evidence test for their stage. The gap between the two numbers is the risk in the book, expressed in dollars. Our pipeline coverage entry has the formula and the evidence-weighted version, and the pipeline review template stages every deal by buyer evidence each week.
Close dates are the loudest risk signal and the least informative
A close date that moves is the risk signal every tool and every manager notices first. Avoma lists a passed or changed close date as one of its five deal risks, and it is right to.
The trouble is that a moved date says something went wrong without saying what. Many slipping dates trace to row six of the table: the date was the rep's from the start, so nothing on the buyer's side was ever going to hold it. Others trace to rows one and four, where an approval step nobody mapped takes longer than the rep guessed. The coaching move is the same each time: ask what is on the buyer's calendar and who put it there, then plan the conversation that gets a buyer-owned date. How to coach reps on close dates has the questions and the call.
Qualifying out is part of managing risk
Qualifying out means removing a deal from the pipeline when the evidence says it will not close in a time frame that matters. It costs a team very little, and reps tend to resist it.
The rule we use is simple: drop a deal when the buyer has declined a specific, low-cost next step twice. One declined call about a number the buyer gave you can be bad timing. The second is an answer. A rep carrying an at-risk deal for a quarter spends their best hours on the buyer least likely to act, and the forecast carries a number nobody believes. How to qualify out a deal covers the conversation with the buyer, the note in the CRM and the coaching conversation with the rep who does not want to let go.
The forecast call is where risk gets priced
The forecast call is the weekly meeting in which a manager and their reps, or a leader and their managers, agree on the number for the period and the deals behind it. It is where deal risk turns into a commit, a best case or a cut.
Most forecast calls price risk by asking the rep how confident they feel. A better call asks, for each committed deal, for the six rows of the table answered with something the buyer said or did. Deals that fail a row move out of commit until the conversation that fills the gap has happened. How to run a forecast call gives the procedure, the forecast call agenda template gives a copyable agenda, and how to improve forecast accuracy covers the habits that move the number over a quarter.
From flag to fix: coaching the at-risk deal on a call
Coaching an at-risk deal means a manager and a rep get on a call, find the missing piece of evidence and plan the buyer conversation that gets it. It takes about 20 minutes, and it ends with an opening line said out loud and a date.
- 01Date the risk3 minutes, what the buyer last did
- 02Walk the table7 minutes, the six rows until one fails
- 03Plan the conversation7 minutes, the opening line said out loud
- 04Decide3 minutes, a date or drop the deal
The call runs in four parts:
- Date the risk (3 minutes). The rep says what the buyer last did and when, and what is supposed to happen next on the buyer's side.
- Walk the table (7 minutes). The manager goes down the six rows in order and asks for the buyer evidence that rules each one out. The first row the rep cannot answer is the gap.
- Plan the conversation (7 minutes). Who the rep talks to, the opening line in the rep's own words, the buyer's likely answer and the rep's reply. The rep says the opening line out loud, and the manager listens for anything that sounds like a check-in.
- Decide (3 minutes). The conversation happens by a date, or the deal is dropped under the twice-declined rule.
A flag changes nothing until someone has the buyer conversation, and the coaching call is where it gets planned. A coaching call is only worth having if it ends in a better buyer conversation. The deal coaching guide covers the coaching side in depth, and how to coach a stalled deal applies it to deals that have gone quiet.
A worked example: one deal, one flag, one conversation
This example is illustrative. The company, people and figures are invented.
The flag. Dana manages six reps. On Monday her forecasting tool marks a $90,000 deal of Sam's as high risk: engagement has dropped, only one contact has been active for three weeks, and the close date has moved once. Sam has it in commit for this quarter.
The call. Dana and Sam get on a 20-minute call. Sam dates the risk: the last buyer action was 16 days ago, when the champion, an operations director called Lee, said he would "take it to the leadership team." Nothing is booked.
Dana walks the table. Row 1, economic buyer: has the person who signs been on a call? Sam knows the VP of Finance will sign, but the VP has never been on a call. Row one fails, so Dana stops there, as the table says.
The diagnosis. Sam does have a cost: Lee said on the second call that manual reconciliation costs his team "about two days a week across four people." But Lee has never had to make that case out loud, so the VP of Finance has never heard it. The tool saw low engagement and a single thread; both are true, and neither tells Sam what to do.
The conversation. Sam plans a call with Lee. Opening line: "Before this goes to your VP of Finance, can we spend 20 minutes on the two-days-a-week number, so she hears it the way you described it to me? And when is she likely to look at it?" Likely answer: "I will handle it, she is busy." Sam's reply: "Then let's write the two lines she will ask about, and I will be on standby if she has questions." The exit is a date for the VP to see the case. Sam says the opening line out loud to Dana before the call ends.
The decision. Sam calls Lee by Wednesday. The deal moves from commit to best case until the VP has seen the case. If Lee declines the 20 minutes twice, the deal comes out of the quarter.
Sam now knows who to call, what to say and where the deal sits in the quarter. The score will catch up.
Doing this for every deal in the book
The procedure above works for one deal. A manager with eight reps and 80 open deals cannot run the table on every deal every week, and that is the real constraint on deal risk: there are never enough good managers to coach every rep on every deal.
Replicate Labs is built for that gap. The expert flags a deal at risk in Slack, email or the Workbench, with what the buyer has and has not said, and Bryce's Monday read tells the manager which deals need attention this week. It also sees which deals are in trouble.
The difference is what happens next: the rep clicks into a call with Caty, talks the deal through and leaves with the opening line for the conversation that fills the gap, in the team's own method. Typing is there for when the rep cannot talk. Bryce then coaches the manager on the 1:1 and the forecast. Drive deal execution shows the full workflow. To see it on one deal, take whichever deal in your book went red most recently and talk to us before anyone sends a nudge email.
FAQ
What is deal risk in sales?
Deal risk is the chance that an open opportunity will not close on the terms and date in the forecast. It usually traces to a missing piece of buyer evidence, such as a signer who has never heard the case or a close date the buyer never agreed to.
How do you identify at-risk deals?
Run two tests on each deal. The evidence test asks whether the buyer evidence for the deal's stage exists, such as the economic buyer on a call by the proposal stage. The time test asks whether the buyer's last action is recent and a dated buyer step exists. A deal that fails either is at risk, whatever its score or stage.
What are the most common signs a deal is at risk?
The common signs are a champion who goes quiet, a single contact on every call, no senior buyer involved, late requests for comparisons, and a close date that moves without a buyer reason. Each points to a missing piece of evidence.
What should a manager do when a deal is flagged at risk?
Get on a 20-minute call with the rep. Date the risk, walk the six evidence rows in order until one fails, plan the buyer conversation that fills that gap, and have the rep say the opening line out loud. End with a date for that conversation or a decision to drop the deal.
How does deal risk affect the sales forecast?
At-risk deals sitting in commit are a common reason forecasts miss. A forecast call that asks for buyer evidence on each committed deal, and moves deals that fail a row into best case until the gap is filled, ties the number to what buyers have done.
When should you remove an at-risk deal from the pipeline?
Remove it when the buyer has declined a specific, low-cost next step twice. One declined call can be bad timing; two is an answer. Take the deal out of the forecast, send a short note that leaves the door open, and record the reason in the CRM so the next rep knows why it stopped.