How do you run a deal review that actually moves deals forward in 2027?

What is the difference between a deal review and a forecast call in 2027?

A forecast call answers "will it close and when," rolled up across a team so leadership can commit a number. A deal review answers "what has to be true for this specific deal to advance, and what are we doing about the gaps." Conflating the two is the single most common structural mistake, because a forecast call rewards optimistic certainty ("commit, 90%") while a deal review rewards honest exposure of weakness. When you run them as one meeting, reps learn to hide risk to protect their commit, and the review stops surfacing the very problems it exists to fix.

The practical separation is temporal and cultural. Forecast happens on a fixed cadence with the full team, is metrics-first, and is deliberately fast. Deal reviews are smaller — the rep, their manager, and maybe a specialist or exec sponsor — and they are slow on purpose, going deep on two or three deals rather than skimming twenty. In a healthy 2027 org, revenue intelligence handles the roll-up automatically, so the forecast call shrinks to exception handling and the freed hours flow into real deal work. If you want the qualification backbone that makes reviews rigorous, the MEDDPICC qualification framework is the connective tissue between the two rituals.

How do you prepare for a deal review so the meeting is not wasted?

Preparation is where 80% of a deal review's value is created or destroyed. The rep should arrive with a completed qualification snapshot, a stated point of view on the deal's true probability, and a specific ask for the room — coaching, an exec intro, a pricing exception, a decision on whether to walk. A review where the manager is discovering the deal for the first time degrades into fact-finding, and fact-finding is exactly what the automated systems already did. The manager's prep is different but equally mandatory: pre-read the AI call summaries, note the sentiment trend, and flag the two questions they most want answered before they sit down.

The forcing function is a lightweight pre-read template the rep fills in and the manager reviews asynchronously. It captures the compelling event, the economic buyer, the decision criteria and process, the identified pain, the champion and their power, competition, and the paper process — plus a single "biggest risk" line the rep must write in plain language. If the rep cannot name the biggest risk, that is itself the finding. This asynchronous prep is what lets the live meeting be a working session instead of a briefing, and it is the highest-leverage change most teams can make this year.

The diagram above shows the whole point: the live meeting is a narrow slot bracketed by heavy prep and hard decisions. Skip the prep and the "working session" collapses back into the rep reading the CRM aloud while the manager nods.

What questions should a manager actually ask in a deal review?

The best deal-review questions are diagnostic, not rhetorical, and they target the gaps between what the rep believes and what the buyer has actually done. Instead of "how confident are you," ask "what has the economic buyer personally done in the last two weeks that proves this is a priority?" Instead of "when will it close," ask "walk me through the exact steps from today to signature, and who owns each one." These questions convert vague optimism into verifiable evidence, and they expose single-threaded deals, phantom champions, and compelling events that were never real.

A disciplined manager rotates through a small set of proven angles: the multi-threading check (who besides your champion wants this), the negative-consequence test (what happens to the buyer if they do nothing), the competitive frame (why us over the alternative and over status quo), and the paper-process reality check (has anyone touched procurement or legal). The goal is to find the one weakest link and coach the rep to strengthen it before the next call. Research on win rates consistently shows that deals with a validated economic buyer and multiple stakeholders close at materially higher rates than single-threaded ones, which is why those two questions earn their place every single review. For structuring the actual conversation with the buyer, pair this with the discovery call framework so the answers you demand in review are answers the rep can realistically go get.

The failure mode to avoid is the manager who uses questions to perform expertise rather than to help. If every question makes the rep feel smaller and no question produces a next action, the review is theater. Great questions end with the rep saying "I don't know — and here's how I'll find out by Thursday."

How do you turn a deal review into concrete next steps that advance the deal?

Every deal review must end with a short, written list of owned actions with dates and names, tied directly to the risks surfaced in the room. "Follow up with the buyer" is not a next step; "email the CFO a one-page ROI model by Wednesday and secure a 20-minute review by Friday" is. The discipline is to convert each identified gap into a specific advancement action and to assign it to a person — usually the rep, sometimes the manager or an exec sponsor. Without this, reviews generate insight that evaporates the moment everyone leaves the room.

The strongest teams close the loop by making next steps visible and revisited. The actions land in the CRM or the deal workspace, the next review opens by checking whether last review's actions moved the deal, and repeated failure to advance triggers a kill-or-escalate decision. This creates accountability without micromanagement: the system, not the manager's memory, tracks whether commitments turned into motion. The map below shows how a single review turns exposed risk into governed advancement.

Notice the "stalled twice" branch. A review process that only ever advances deals and never kills them is not qualifying — it is hoarding pipeline. Killing a dead deal fast is a win because it returns the rep's most finite asset, selling time, to deals that can actually close.

How do AI and revenue intelligence change deal reviews in 2027?

By 2027, revenue-intelligence tooling has moved the baseline: call recording, transcription, sentiment scoring, and automatic stakeholder mapping are table stakes, and the CRM is largely populated by AI rather than by rep data entry. This changes the review's center of gravity. The machine handles "what happened" — who was on the call, what was said, whether the deal has gone quiet — and flags anomalies like a champion who stopped replying or a stalled paper process. That frees the human meeting to focus entirely on "what should we do," which is still stubbornly a judgment problem.

The risk is over-trusting the model. AI deal scores are pattern-matched probabilities, and they are wrong in exactly the interesting cases — the strategic deal that looks weak on activity but has a committed exec sponsor, or the deal that scores hot because of engagement that is really the buyer doing diligence to renew a competitor. A good 2027 manager treats the AI score as one input to interrogate, not a verdict to accept, and explicitly asks "where does the model disagree with your gut, and who's right?" The teams getting real leverage use AI to eliminate the reporting portion of reviews entirely and reinvest 100% of that time in coaching and buyer strategy. For the operational side of wiring these signals into your process, see how to build a RevOps deal desk.

There is also a cultural shift. When every call is captured and summarized, reps cannot bluff a review with a rosy verbal recap, and managers cannot coach from anecdote. The evidence is on the table. That transparency is uncomfortable at first and enormously productive once the team adapts, because arguments move from "I think the champion is strong" to "the champion has not been on a call in 18 days — let's fix that."

How often should you run deal reviews and which deals qualify?

Not every deal deserves a review, and reviewing all of them equally is a fast way to make the ritual worthless. Segment by stakes: the largest strategic deals warrant a deep, recurring review — often weekly or biweekly in late stage — while the mid-market bulk of pipeline gets a lighter, exception-based look triggered by risk signals like slippage, going quiet, or a stalled stage. Trying to give every deal the full treatment dilutes attention and burns out both rep and manager. Cadence follows deal value and deal risk, not a rigid calendar applied to everything.

The qualifying filter should be explicit and shared. A deal earns a full review when it crosses a revenue threshold, when it has slipped a close date, when the AI flags a momentum drop, or when the rep requests help. Everything else stays in the automated forecast layer until it trips a trigger. This tiering keeps the reviews that happen genuinely high-value and prevents the calendar from filling with meetings about deals that are either too small to matter or too healthy to need help. Pair the cadence with a clean stage definition so "advanced a stage" means the same thing to everyone; loose stage definitions are the quiet reason many reviews argue in circles about whether a deal actually moved.

Related questions

What is the biggest mistake managers make in deal reviews?

Treating the review as a forecast interrogation. When the meeting rewards confident commits over honest risk exposure, reps hide problems to protect their number, and the review stops surfacing the issues it exists to catch. The fix is separating deal reviews from forecast calls entirely.

How long should a deal review be?

Go deep, not wide. Spend 20 to 30 focused minutes on two or three real deals rather than five minutes each across twenty. Depth surfaces the weak link; breadth just produces a status recital the CRM already contains. Let automated forecasting handle the roll-up.

Should reps run their own deal reviews?

Yes — self-review is a core skill. Reps should complete the qualification snapshot and name the biggest risk before any manager sees it. The manager's job is to pressure-test that self-assessment, not to discover the deal from scratch. Reps who cannot self-qualify are not yet ready to forecast.

What framework works best for deal reviews in 2027?

MEDDPICC remains the dominant enterprise qualification backbone because it maps directly to the questions a review must answer: economic buyer, decision process, metrics, champion, and paper process. Lighter deals can use a trimmed version, but the underlying discipline of evidence over optimism is universal.

How do you kill a deal without demoralizing the rep?

Frame the kill as reclaiming selling time, not as failure. A dead deal identified in week 6 saves the rep three months of chasing. Celebrate fast, honest disqualification the same way you celebrate a close — both protect the rep's finite capacity to sell.

FAQ

What should a rep bring to a deal review? A completed qualification snapshot, a stated true-probability point of view, the single biggest risk written in plain language, and a specific ask for the room — coaching, an intro, an exception, or a kill decision. Arriving without these turns the review into fact-finding the systems already did.

How is a 2027 deal review different from 2020? Revenue intelligence now auto-captures and summarizes every interaction, so the "what happened" portion is eliminated. The meeting focuses entirely on judgment and next actions. Reps can no longer bluff with rosy recaps, and managers can no longer coach from anecdote — the evidence is on the table.

Who should attend a deal review? Keep it small: the rep, their manager, and occasionally a specialist or executive sponsor for strategic deals. Large audiences make reps defensive and slow the working session. Forecast roll-ups are where the full team belongs, not deal reviews.

How do you measure if deal reviews are working? Track stage-advancement rate after reviews, next-step completion rate, and time-to-kill on dead deals. Reviews that generate written, owned actions and then verify them next session are working. Reviews with no measurable change in deal motion are theater, regardless of how insightful they feel.

What is the role of AI deal scores in a review? Treat scores as one input to interrogate, never a verdict. Explicitly ask where the model disagrees with the rep's judgment and who is right. AI is excellent at flagging anomalies like a gone-quiet champion and weak at strategic deals with committed sponsors but low activity.

How do you handle a rep who inflates every deal? Anchor on evidence, not opinion. When every claim must be backed by a buyer action from the last two weeks, inflation collapses on its own. The captured-call record removes the ability to bluff, and repeated gaps between claim and evidence become a coaching conversation about honesty and self-qualification.

Should deal reviews and pipeline reviews be the same meeting? No. Pipeline review looks at coverage and health across the whole book; deal review goes deep on specific opportunities. Combining them means neither gets done well — you skim everything and interrogate nothing. Separate the cadences and give each its own purpose.

How do you keep deal reviews from becoming micromanagement? Make the system, not the manager's memory, track commitments. Log next steps in the deal workspace, revisit them at the top of the next review, and let repeated stalls trigger escalation automatically. Accountability lives in the process, freeing the manager to coach rather than police.

Sources

• Gartner Sales Research — Buying and Deal Process
• Forrester — B2B Revenue Operations Research
• MEDDICC / MEDDPICC Official Framework
• Gong Labs — Revenue Intelligence Research
• Salesforce — State of Sales Report
• Harvard Business Review — Sales Management
• Korn Ferry (formerly CSO Insights) — Sales Performance Research
• Winning by Design — Revenue Architecture
• RevOps Co-op — Community Resources

Related on PULSE

• MEDDPICC qualification framework
• The discovery call framework
• How to build a RevOps deal desk
• Reading pipeline health signals