Flagged, Then Left Waiting: Revenue Teams See Deal Risk and Take Days to Act

Ask a revenue leader whether their team can spot a deal going sideways and the answer is almost always yes. Conversation intelligence flags a competitor mention, an engagement score drops, a champion goes quiet. Ask how long it takes for someone to do something about it, and the answer gets less comfortable. With fourth-quarter close weeks away, new research suggests the bottleneck in most pipelines is no longer detection. It is the days that pass between a warning appearing and a person acting on it.

The Signal Arrives, Then Waits for the Calendar

The Revenue Execution 2026 survey from Everstage, fielded independently by Emporia Research in July and August among 195 CROs, VPs of Sales and VPs of Revenue Operations at mid-market and enterprise SaaS and AI-native companies, sums up its findings bluntly: every revenue team has AI now, and most still cannot act on it. When a deal shows risk, 45% of teams take four or more days to respond. That is close to a full business week for a competitor to get in front of the buyer, or for a stalled approval to become a pushed close date.

The reason is structural. Some 57% of respondents say they first learn a deal is at risk in a pipeline review or from a rep, and only 17% first hear about it from an automated alert. In other words, reviews still run on the calendar rather than on the signal. A risk that surfaces on a Tuesday waits for the Monday forecast call, and then waits again for someone to own the next step.

The data is not missing, either. Eight in ten leaders say they make revenue decisions before all the data arrives, yet more than half say the information already existed somewhere in their stack. Everstage frames that as a routing problem rather than a visibility one, and 74% of respondents say the lag between knowing and acting is costing them real revenue. On its Revenue Execution Week page, the company adds that 51% of leaders say their AI surfaces an insight and stops there.

Detection has been automated. Response is still waiting for the next meeting on the calendar.

Where the Days Actually Go

Forecasting is the clearest example. Building the weekly forecast takes two hours every week, and seven in eight leaders in the Everstage survey still build it by hand. Those hours go to assembling a picture of the pipeline, not to changing it. The same pattern shows up in tooling more broadly: in the Salesloft 2026 US Revenue Benchmark of 500 sales and revenue leaders, every respondent said their organization uses AI somewhere in the revenue process, but only 20.6% described their AI strategy as production-ready with measurable outcomes, and 55.6% said CRM information is based mostly on subjective seller reporting.

Pricing is the second choke point, and arguably the most expensive. Everstage reports that 47% of revenue organizations take three or more days to approve a pricing exception when the buyer is ready to sign. A late-stage deal is at its most fragile in exactly that window. The buyer has a signature authority lined up and a budget cycle to hit, and the seller is waiting on a deal desk, a finance approver or a spreadsheet that has to be rebuilt by hand.

Then there is the people layer. In smaller companies, the operating rhythm that would catch a slipping deal often does not exist at all. The eighth edition of the Sales Xceleration State of Sales report, built on assessments from 5,263 leaders at companies with $1 million to $200 million in revenue, found that 66% have no documented sales process, 88% lack clarity on what must happen at each stage to advance a deal, 82% do not forecast future sales, and 66% do not hold weekly one-on-ones between leaders and reps. Without stage definitions, a risk signal has nothing to be measured against, and without a regular coaching conversation, nobody is assigned to act on it.

Closing the Gap Means Redesigning the Workflow

The organizations getting value from AI are not the ones with the most alerts. In a Gartner survey of 227 chief sales officers, sales organizations that give sellers AI-enabled next best actions were 2.6 times more likely to achieve commercial growth, and those that prioritize upskilling sellers on AI were 2.4 times more likely to achieve strong revenue growth. The difference between an insight and a next best action is ownership: one tells you something is wrong, the other tells a specific person what to do about it, and when.

"The most effective sales organizations are not simply layering AI onto existing ways of working," said Greg Hessong, Senior Director Analyst in the Gartner Sales practice. That is the core of the problem the Everstage data describes. Teams added detection on top of a weekly review cadence, a manual forecast and a multi-day approval chain, and then wondered why faster signals did not produce faster outcomes.

The CRO Playbook for Shrinking Time to Action

• Measure response time, not just detection. Track the hours between a risk signal firing and a logged action on the deal, by segment and by manager. If you cannot report it, you cannot shorten it.
• Give every alert an owner and a deadline. Route deal-risk signals to a named person with an expected response window, rather than to a dashboard that waits for the weekly review.
• Move deal reviews from the calendar to the trigger. Keep the weekly forecast call, but convene a short, focused review whenever a late-stage deal crosses a defined risk threshold.
• Pre-approve the common exceptions. Set discount guardrails and approval tiers inside the quoting workflow so routine pricing requests clear in hours, and reserve escalation for the genuinely unusual deal.
• Define what each stage requires. Written exit criteria for every stage give risk signals something to be measured against and make it obvious when a deal has stopped advancing.
• Coach managers to act on signals, not just read them. Build weekly one-on-ones around the deals that flagged risk since the last meeting, and train managers on the specific plays that follow each warning.

Revenue teams spent the last two years buying the ability to see trouble coming, and by most accounts that investment worked. The next gain will not come from another alert. It will come from collapsing the four days between the warning and the response into four hours, because in a tight fourth quarter, the deal that is flagged and left waiting is the deal that quietly slips into next year.