How do you coach a rep who skips discovery and jumps to the demo?

The outcome you should expect

Managers usually want one thing from this coaching: the rep stops burning first meetings on generic product tours. That is the right goal, but it is not the first thing that moves, and if you watch the wrong number you will conclude the coaching failed while it is actually working.

Here is the honest sequencing. In weeks one through three, the visible change is small and mechanical — the rep starts asking two or three more questions before sharing their screen, often stiffly, often reading from a card. It sounds worse before it sounds better, the same way a golfer's swing degrades when a coach changes their grip. Do not panic at this stage and do not let the rep panic either. Tell them in advance that the first ten calls will feel clumsy, because that prediction, made ahead of time, is what keeps them from abandoning the new motion the first time a call goes sideways.

By weeks four through six, the second-order effect appears: demos get shorter and more specific. The rep who used to run a forty-minute feature parade starts running a fifteen-minute demonstration of two or three things tied to something the buyer actually said out loud. Buyers respond to this differently — you will hear more questions from the buyer during the demo instead of polite silence, and questions during a demo are the single best qualitative signal that the rep earned the right to be there.

By weeks eight through twelve, the pipeline math changes. Fewer demos get booked, and that is a feature, not a bug — reps who qualify properly disqualify more. The number that should rise is demo-to-next-step conversion, because an earned demo produces an agreed follow-up and a thrown demo produces "let me circulate this internally and get back to you," which is a polite obituary. Total closed revenue is the last thing to move, typically one full sales cycle behind the behavior change, which for most B2B teams means a quarter or more.

Set expectations with your own leadership accordingly. If you tell your VP that fixing discovery will lift bookings this month, you have set yourself up to abandon a working intervention at week five. Tell them instead that demo volume will dip, demo quality will rise, and conversion downstream of the demo will move first. That framing survives contact with a mid-quarter pipeline review.

There is also an outcome you should expect that nobody mentions: some reps will not change, and the coaching will have been valuable anyway because it produced a clean, documented answer to the question of whether they can. A month of structured coaching with recordings, scorecards, and role-plays turns a vague "he's not doing discovery" complaint into evidence. That evidence is what makes the next conversation — the performance one — fair and fast rather than drawn out and contested.

What drives that outcome

The behavior has four root causes and they need four different interventions. Getting this wrong is the most common failure in the whole exercise, because a manager who applies drills to a belief problem gets compliance that evaporates the moment supervision stops.

Skill gap. The rep does not know how to run discovery. They ask one surface question, hear a keyword they recognize, and reach for the demo because the demo is the only motion they are actually fluent in. This is the most common cause and the most fixable. Signs: they fumble in a role-play, their questions are all closed-ended, and they cannot restate the buyer's problem in the buyer's words after the call.

Belief gap. The rep can run discovery — you have watched them do it off-stage — but they believe questioning stalls a hot buyer, or they believe the product is compelling enough that setup is unnecessary. This is a conviction problem and no checklist touches it. The only lever that works is evidence from their own deals: pull two won deals with strong discovery and two lost deals with a premature demo and let them narrate the difference.

Knowledge gap. The rep is willing and capable but does not know which questions matter for your specific buyer. A rep who moved from selling a self-serve tool to selling a platform with a six-person buying committee will default to a tour simply because nobody handed them the question bank. Fix this with a written set of questions mapped to your qualification framework — MEDDIC, SPIN, SPICED, whatever your team already uses — not with more encouragement.

System problem. If your comp plan pays on demos booked, if your dashboard celebrates demo volume, or if your SDR-to-AE handoff SLA requires a demo scheduled within forty-eight hours, you are paying the rep to skip discovery. This is the cause managers are least likely to check because it implicates them. Audit it first. A rep responding rationally to a broken incentive is not a coaching case.

The bottom-right branch is the one managers avoid. If you have diagnosed skill, supplied the questions, run the drills weekly, shown the evidence, and the rep still will not change after a month, you do not have a coaching problem anymore. A fifth role-play is avoidance dressed up as patience.

One more driver worth naming, because it sits upstream of the rep entirely: lead quality. A rep working a list of unqualified inbound clicks learns, correctly, that most conversations are not worth a discovery investment. They skip to the demo because the demo is a fast disqualifier. If your marketing-sourced meetings convert at a fraction of your outbound-sourced ones, the rep's shortcut is a rational adaptation to a RevOps problem, and no amount of 1:1 coaching will out-argue their lived experience. Fix the routing and the scoring, then coach.

Benchmarks and realistic ranges

You need numbers to know whether coaching is working, and the useful ones are all leading indicators. Lagging quota attainment tells you nothing for a full cycle. Build a small set and track the trend per rep rather than chasing an industry average that may not apply to your motion.

Discovery-to-demo ratio. What share of demos are preceded by a genuine discovery call — not a fifteen-minute intro, an actual scoped conversation? Most teams that have never measured this discover the number is worse than they assumed. Set the target at your own historical top-quartile rate rather than an arbitrary hundred percent, because some inbound motions legitimately compress discovery and demo into one call. If yours does, the standard is not "separate calls," it is "quantified pain before screen-share," which is measurable inside a single meeting.

Discovery scorecard average. Grade every reviewed call on three dimensions, one to five each. *Pain clarity*: can the rep state the buyer's top problem in one sentence using the buyer's own language? *Impact quantification*: did they ask what the problem costs in time, money, or headcount, and get an actual number? *Next-step logic*: did they set a specific follow-up that ties to the pain, not a generic "let's touch base"? Fifteen points available. Pick a floor — many teams land somewhere around twelve — and make that the gate for booking a demo. The gate matters less than the fact that it is written down and consistently applied.

Demo-to-next-step conversion. This is the number that moves first and the one I would put on the wall. Earned demos advance; thrown demos die. If a rep runs eight demos and six produce a mutually agreed next step with a date, discovery is working. If six of eight produce "I'll take this to the team," it is not.

Talk-to-listen ratio on discovery calls. Conversation-intelligence platforms surface this automatically. A rep who jumps to the demo is almost always over-talking on the calls that precede it. You do not need a precise industry benchmark to use this — compare the rep against your own top performers on your own calls, which is a far more defensible standard than a blog-post number.

Multi-threading rate before demo. How many distinct stakeholders has the rep engaged before showing product? In a committee purchase, a single-threaded demo is a guess about what five other people care about. Track the count and watch whether it rises as discovery quality rises — the two move together, because a rep who asks about decision process learns there are other people.

Forecast slippage on demoed deals. Premature demos inflate the pipeline with deals that stall in the middle stages. If your stage-two-to-stage-three conversion is healthy but stage-three-to-close is dismal, the demo is manufacturing false progress. Pull this from your CRM or forecasting tool and look at it by rep, not just in aggregate — one rep stuffing the pipeline can hide behind a healthy team number.

On timeframes: skill-driven cases commonly shift within three to four weeks of weekly drills and call reviews, because you are building a motion, and motions build fast with repetition. Belief-driven cases take longer, frequently two to three months, because you are changing a conviction the rep formed from experience and one conversation does not overturn it. Treat a month with zero movement under proper coaching as diagnostic information rather than a reason to coach harder.

A caution on all of these: do not stack six new metrics on a rep at once. Pick two — scorecard average and demo-to-next-step is a good pair — and add others only once those are stable. Metric overload produces gaming, and a rep who games a discovery scorecard by asking scripted questions without listening to the answers has learned something worse than the original habit.

Risks, edge cases, and failure modes

The buyer genuinely wants a demo first. Some do, and they are not wrong. A technical evaluator who has already read your docs and shortlisted three vendors may reasonably want to see the product before investing in a conversation. Coaching a rep to stonewall this buyer costs you deals. The correct move is a partial concession: show something small and specific quickly, then use what the buyer reacts to as discovery material. "Here's the piece most teams like yours start with — tell me if that's even the right area" turns a demo into a diagnostic. Reps need permission to run this variant or they will feel trapped between your rule and the buyer's request.

Compliance masquerading as change. The rep asks the questions while you are watching and reverts the moment you are not. This is the most common failure and the reason spot-checking beats scheduled review. Pull an unannounced call occasionally, not to catch them, but because a metric that only holds under observation is not a metric. If scorecard averages are strong on submitted calls and weak on random pulls, you have a will problem wearing a skill problem's clothes.

Rescuing the deal instead of building the skill. A manager jumps on the call to "show them how," saves the deal, and teaches the rep nothing except that the manager will bail them out. Every rescue you run is a rep that did not learn. If you must join a call for revenue reasons, debrief it afterward as a teaching artifact and name explicitly what you did and why, or the value evaporates.

Coaching everyone identically. The skill-gap rep needs drills. The belief-gap rep needs evidence and a different conversation entirely. Running the same 30/60/90 program across a whole team wastes the time of the reps who did not need it and under-serves the ones who need something else. Diagnose per person.

The new-hire edge case. A rep in their first ninety days who skips discovery is not exhibiting a bad habit — they are exhibiting an untrained one, and the intervention is onboarding, not coaching. Do not put a thirty-day rep on a performance-flavored improvement loop for a skill nobody taught them. Conversely, a five-year veteran who skips discovery has usually been rewarded for it somewhere, which points you at the system.

The product that genuinely demos well. In some categories — visual tools, design software, anything with an obvious wow moment — the demo really does create desire faster than questions do. If your best reps consistently lead with a short demo and still win, your "earn the demo" rule may be fighting a motion that works. Verify against your own win data before installing a rule that contradicts it. The underlying principle survives: the rep must know what problem they are solving before they propose a solution. Whether that knowledge is acquired before or during the demo is a sequencing choice, not a moral one.

Over-correction into interrogation. A rep who takes the coaching too literally turns discovery into a deposition — twenty questions, no reciprocity, buyer exhausted. Discovery is a conversation with insight traded back, not a form being filled out. Watch for the rep who scores well on your scorecard while buyers stop responding to their follow-ups; that pattern is the tell.

Downstream damage from the fix. When reps start disqualifying properly, top-of-funnel metrics that other teams are measured on get worse. Your SDR team's meeting-to-opportunity rate drops because AEs are rejecting more. If you have not warned marketing and sales development in advance, you will spend week six in a meeting defending a change that is working. Bring RevOps in early so the funnel definitions and the dashboards move together with the behavior.

A practical rollout plan

Run this as a structured ninety-day loop with a specific weekly rhythm. One good conversation changes nothing; the cadence is the intervention.

Before day one — audit the system. Look at the comp plan, the dashboards, the SDR handoff SLA, and the CRM stage definitions. If anything there rewards demo volume or requires a demo booked within a fixed window, fix it or you are coaching against your own incentives. Also pull two won and two lost deals for the rep so you have their own evidence ready.

Days one to thirty — build the motion. Open with a focused 1:1 structured around goal, reality, options, and commitment. Do not open with "you're skipping discovery." Open with a call they are proud of, then play a recording and stop at the flip point — the exact second they moved to screen-share. Ask: walk me through what happened here. Then: what did we know about their actual problem at that moment? Most reps go quiet. That silence is the coaching, because they discover the gap themselves rather than being told about it.

Then co-create the fix instead of dictating it. Ask what they would ask if they could rewind, build their list, add yours. Hand them redirect language they can actually say out loud without sounding evasive: "Absolutely — and I want to show you the right part, not a generic tour. Ninety seconds: what's the one thing that, if this fixes it, makes this an easy yes?" A second variant for the buyer who pushes back: "I can show you the product in thirty seconds, but I'd rather show you your version of it. Five minutes on your situation and then I'll tailor it." Close by having the rep state their own rule — *I don't demo until I've quantified one painful metric and confirmed who else decides* — and write it down. Owned rules survive; imposed rules do not.

Weekly during this phase: one recorded call reviewed against the scorecard, two role-plays on the redirect line, and a daily submission of one call with the flip point timestamped. The role-plays matter more than the reviews. Run the redirect five times back-to-back until it is muscle memory rather than a line being read.

Days thirty-one to sixty — transfer ownership. The rep self-scores their own calls before the 1:1 and you compare notes; the gap between their score and yours is the most useful diagnostic you will get all quarter. Drop recordings to two a week. Shift to live-deal coaching: pick an open opportunity and ask what discovery is still missing on it. A qualification-grid gap-fill works well here — empty cells for economic buyer, decision criteria, or quantified pain are objective, visible, and hard to argue with, which beats a subjective debate about whether discovery was "good enough."

Days sixty-one to ninety — prove durability. Move to spot-checks rather than surveillance. Have the rep teach discovery to a peer in a team session, because teaching cements a skill faster than being taught again. Raise the bar to multi-threaded discovery — the committee, not just the champion — and to harder objection handling. If the behavior holds under reduced observation, it is real.

Scaling it past one rep. If three or more people on the team share this habit, stop running individual coaching and treat it as a team-level design problem. Build the question bank once, add a required discovery-summary field on the opportunity record so the CRM enforces what your rule asks for, and put a call-review block on the team calendar so the practice is scheduled rather than squeezed. This is where RevOps earns its keep: the difference between a manager nagging six reps and a process that makes the right behavior the path of least resistance is entirely a systems difference, and systems outlast managers.

Two things to avoid in the rollout. First, do not make the scorecard punitive in week one — a rubric introduced as a judgment gets gamed, while the same rubric introduced as a shared diagnostic gets used. Second, do not skip the written rule. The single highest-leverage artifact in this whole program is one sentence the rep wrote themselves about what must be true before they share their screen, because it is the only part of the system that operates when you are not in the room.

Related questions

Should the rep ever demo on the first call?

Yes, when the buyer is late-stage, technically literate, and comparing shortlisted vendors. The rule is not "never demo early," it is "never demo blind." A short, targeted demo used as a diagnostic — show one thing, ask what they react to — satisfies both.

How do you coach this remotely when you can't sit in on calls?

Conversation-intelligence recordings make remote coaching easier than in-person, not harder. You can review the exact flip point asynchronously, timestamp it, and send it back. Replace ride-alongs with recorded-call review plus a weekly video role-play block.

What if the rep hits quota anyway?

Coach the habit, not the outcome. A rep hitting quota with poor discovery is usually riding strong inbound or a small number of large deals, and the pattern breaks the moment either changes. Frame it as durability rather than deficiency.

Does this apply to sales engineers too?

Often more so. An SE whose entire job is the demo has less natural incentive to qualify, yet an SE who asks two grounding questions before opening the product changes the meeting. Coach them on the same earn-the-demo principle with a technical question set.

How does discovery coaching change for a renewal or expansion motion?

The questions shift from problem discovery to usage and outcome discovery — what changed since last year, what did they actually adopt, who owns the budget now. Skipping straight to a new-feature demo in a renewal is the same error with higher churn stakes.

FAQ

Why do reps skip discovery and jump to the demo in the first place?

Usually one of four causes: a skill gap where they cannot actually run discovery, a belief gap where they think questioning stalls a hot buyer, a knowledge gap where they do not know which questions matter for your buyer, or a system problem where comp plans and dashboards reward demo volume. Diagnose which one before correcting. The wrong fix produces temporary compliance and no durable change, and you will be having the same conversation in six weeks.

What is the single best line to redirect a buyer who says "just show me"?

"Absolutely — and I want to show you the right part, not a generic tour. Ninety seconds: what's the one thing that, if this fixes it, makes this an easy yes?" It honors the request rather than refusing it, reframes discovery as precision rather than gatekeeping, and gives the rep the one piece of pain they need to make the demo land. Rehearse it until it comes out conversationally.

How long should this take to fix?

Plan a ninety-day cycle. Skill-driven cases often shift within three to four weeks of weekly drills and recorded-call reviews because you are building a motion, and motions respond quickly to repetition. Belief-driven cases run longer, commonly two to three months, since you are changing a conviction. Revenue impact trails behavior change by roughly a full sales cycle, so do not judge the program on this quarter's bookings.

When does this stop being a coaching problem and become a performance problem?

When the rep can run discovery cleanly in a role-play but will not live, after you have supplied the question bank, run weekly drills, and shown them the evidence from their own won and lost deals. That combination points at will, not skill, and will does not respond to a fifth role-play. Move to a documented performance conversation — the coaching record you built makes that conversation fair and fast.

Should I use conversation-intelligence tools for this?

They help significantly. Recording platforms let you find the exact moment screen-share started and how much was known before it, which turns a vague complaint into a specific, reviewable clip. Use the tooling to locate the flip point and to track talk-to-listen trends. Keep the human 1:1 for the belief work, which no automated flag will do for you.

How is this different when the buyer is a committee rather than one person?

It matters more. A premature demo to a single stakeholder ignores the several other people who will actually decide, and that stakeholder cannot re-sell your product internally on a tour they only half-understood. Coach the rep to treat multi-threading as part of discovery — who else needs to see this, what does each of them care about — before any demo happens.

Sources

• RAIN Group — Sales Discovery Questions and Best Practices
• MEDDIC Academy — What Is MEDDIC?
• Gong Labs — Sales Research and Call Data
• Harvard Business Review — The New Science of Sales Force Productivity
• Winning by Design — Revenue Architecture Resources
• Sandler — Sales Training Resources
• Richardson Sales Performance — Sales Coaching Insights
• Challenger — Sales and Marketing Research

Related on PULSE

• [How do you coach a rep to tailor the demo to the buyer's pain?](/knowledge/cg0063)
• [How do you coach a rep to give a demo that closes?](/knowledge/cg0061)
• [How do you coach a rep to control the room in a group demo?](/knowledge/cg0069)
• [How do you coach a sales engineer to sell, not just demo?](/knowledge/cg0206)
• [How do you coach reps to set next steps at the end of a demo?](/knowledge/cg0070)
• [How do you coach reps to handle questions during a demo?](/knowledge/cg0064)