How do you coach reps to use MEDDIC in discovery?

The Tuesday deal review that exposes everything

Picture a mid-market AE — call her a two-year tenured rep carrying a $900K number — walking into a Tuesday 1:1 with a deal she's forecast at 80%. The opportunity record is beautiful. Every MEDDIC field in Salesforce is populated. Metrics: "reduce manual reporting time." Economic buyer: "VP of Operations." Decision criteria: "ease of use, integration, price." Decision process: "eval, then legal, then signature." Identify pain: "current process is inefficient." Champion: the ops manager she's been emailing for six weeks.

Now run the drill. Ask her to read you the customer's actual words behind each field. Not her summary — their language, from a call recording or an email she can pull up on the spot.

The Metrics line collapses first. "Reduce manual reporting time" is her phrasing, not theirs. Nobody on the buyer side ever said a number. She thinks it's maybe ten hours a week, but that's an inference she made from a comment about Monday mornings being rough. Economic buyer: she's never spoken to the VP. She knows the name because it was in an org chart on LinkedIn. Decision criteria: those three items came from her own discovery deck, which she walked through on call two, and the ops manager nodded. Decision process is a guess assembled from how her last two deals went at other companies. Identify pain is a paraphrase. And the champion has never once offered to introduce her upward — she's asked twice, and both times got "let me see what makes sense."

This deal is not at 80%. It's a deal with one friendly contact, zero quantified value, no verified buying process, and no access to money. It will slip at least one quarter, probably two, and there's a real chance it never closes at all — it dies quietly to "we decided to revisit next fiscal year," which is what no-decision looks like when it arrives politely.

Here is the part that matters for coaching: the rep did nothing lazy. She filled out every field, she had a dozen conversations, she built a real relationship. The failure was that the CRM asked her for *labels* and she supplied labels. Nobody ever asked her for *evidence*. The gap between "MEDDIC is populated" and "MEDDIC is proven" is the entire coaching job, and almost every underperforming sales org lives inside that gap without knowing it.

The scenario generalizes well beyond one AE. RevOps teams see the same pattern in aggregate when they look at forecast accuracy: qualification fields are 90%+ complete across the pipeline and forecast accuracy is still poor. That combination is diagnostic. High field completeness plus low forecast accuracy means the fields are being filled from the rep's imagination, not the buyer's mouth. Low completeness with low accuracy is a simpler problem — reps aren't doing the work at all, and you fix that with process. The insidious version is the well-behaved pipeline full of confident fiction.

So the coaching unit isn't the acronym. It's the quote. Every MEDDIC element either has a customer utterance behind it or it doesn't exist. Teach that one standard and most of the rest follows, because a rep who knows they'll be asked for the quote on Tuesday starts asking the question on Thursday's discovery call.

How the coaching loop actually works

The mechanism has four moving parts, and they only produce behavior change when all four are connected. Observation feeds diagnosis, diagnosis selects the coaching move, the coaching move gets rehearsed before it's used on a real buyer, and the next live call gets observed. Break any link and the loop degrades into either a lecture or a compliance audit.

Observation. You cannot coach discovery from a CRM record, because the CRM record is the rep's interpretation of the call — which is precisely the thing you're trying to inspect. You need the raw call: a recording, a transcript, or at minimum your own presence on a live one. Conversation intelligence platforms — Gong, Chorus, Clari Copilot, and the recording features now baked into most dialers and video tools — solve the scale problem here. One manager can't listen to forty calls a week, but they can scan transcripts for whether a metric question was ever asked and then listen to the four calls where it wasn't.

Diagnosis. Before you coach anything, classify the gap. There are four root causes and they take opposite responses. *Skill*: the rep doesn't know how to phrase a metric question or test a champion without it landing like an interrogation. *Will*: the rep knows exactly how and skips it, usually because asking about budget authority or the "what if you do nothing" question risks souring a call that feels friendly. *Knowledge*: the rep doesn't understand the buyer's business well enough to translate a symptom into an economic number — they can't get from "reporting is slow" to "four analysts, eight hours each, at a loaded rate" because they've never thought about how that company makes money. *System*: your stages, comp, and CRM don't require MEDDIC, so ignoring it is the rational choice.

Coaching skill when it's a will problem is wasted breath, and the rep learns you can't read them. Coaching will when it's actually a system problem makes you the villain for a process gap you own. Route the symptom to the cause first — it takes ninety seconds and it determines everything that follows.

The coaching move. Run it as a GROW conversation — Goal, Reality, Options, Will — with the deal open on the screen. The structure matters because it makes the rep find the gap instead of receiving a verdict, and self-discovered gaps get fixed.

*Goal:* "Let's take the Acme deal. By the time we're done I want you able to name the economic buyer and quote me one hard number in their words. Fair?"

*Reality:* "Walk me through MEDDIC on Acme. Don't summarize — read me what they actually said." When the rep can't produce a quote, say nothing. Let it sit. Then: "So we don't have that yet. What would you have to ask to get it?"

*Options:* build the question bank together, element by element. For Metrics: "What does this cost you today, in dollars or hours a month?" and "If we fix it, what number does your boss expect to move, and by how much?" For Economic buyer: "Who signs off on a spend this size, and have they bought something like this before?" plus the access test — "Would it make sense to get their view directly so we're not building something they'd reject?" For Decision criteria: "What are the three things this has to do for you to pick us over the alternative or over doing nothing?" and the sharper follow-up, "Did your team write those down, or are we helping shape them?" For Decision process: "Walk me through exactly what happens between your yes and a signature — who, what, how long at each step?" and "What's killed deals like this internally before?" For Identify pain: "What happens if you do nothing for two more quarters?" and "Why now, versus six months ago?" For Champion: the intro test — "Would you be comfortable introducing me to your economic buyer?" A real champion says yes and means it.

*Will:* "Which two are you getting on the next Acme call, and when is that call?" Write it down in front of them. "I'll listen to the recording Thursday."

Rehearsal. Skip this and the loop breaks. A rep who agrees a question is good has not learned to ask it under pressure with a buyer who deflects. Five minutes of role-play where you play the evasive stakeholder does more than thirty minutes of discussion. Make them say the words out loud, badly, twice, before they say them to a customer.

Reinforcement through the system. The final link is structural, and it's where RevOps earns its keep. Make the evidence — not the field — a requirement to advance a stage. A deal cannot move from Discovery to Proposal without a quantified metric, a confirmed pain with a compelling event, and a champion who has passed the intro test. Not three populated text boxes: three pieces of evidence a second person could verify. Validation rules can enforce that fields are non-empty; only a manager or a peer review can enforce that they're true, which is why the stage gate and the weekly review have to exist together.

Numbers worth tracking, and what good looks like

Don't measure coaching by closed-won. Closed-won lags by a full sales cycle — in enterprise, that's six to twelve months, which means you'd learn in Q3 whether your Q1 coaching worked. By then you've changed four things and can't attribute any of it. Measure the leading indicators instead, all of which move within two to four weeks of real coaching.

Metric-capture rate. The percentage of discovery calls where the rep obtained a quantified number in the buyer's words. Conversation intelligence tooling can approximate this from transcripts; a manual sample of five calls a week per rep works fine if you don't have the tooling. Expect this to start low — many teams find that fewer than a third of their discovery calls produce a hard number — and to be the fastest-moving indicator once you start asking for quotes on Tuesdays.

Economic-buyer access rate. The share of deals past qualification where the rep has actually spoken with the person who controls the money, versus deals where they only know the name. This one is binary per deal and easy to audit honestly. It's also the single strongest predictor of whether a forecast is real, because a deal where nobody has met the budget holder is a deal being sold to a spectator.

Champion confirmation rate. Deals with a champion who passed the intro test, divided by all deals with a "champion" logged. The delta between those two numbers is your fiction rate. On teams that have never enforced the test, it's common for a large majority of logged champions to be coaches or friendlies — someone who likes the rep and shares information, but has no power and no personal stake in the outcome.

Stage-gate evidence completeness. Not field completeness — evidence completeness. Sample ten deals a month at random, pull the quotes, and score how many fields are backed by something a customer actually said. The gap between the CRM's reported completeness and your sampled completeness is the number to shrink.

Slippage and no-decision rate. As qualification tightens, two things happen and they look contradictory. Pushed close dates should fall, and your loss-to-no-decision rate should fall — but early on your *disqualification* rate should rise sharply, because reps start killing deals in month two instead of nursing them for three quarters. Warn your leadership before this happens. A pipeline that shrinks 20% while forecast accuracy improves is a win that looks like a disaster on a dashboard nobody prepped for.

Win rate on evidence-complete versus incomplete deals. This is the proof point that ends the argument. Split your closed deals from the last four quarters by whether they had a verified metric, verified economic-buyer access, and a tested champion at the point they entered Proposal. Compare win rates. Nearly every team that runs this analysis finds a meaningful spread, and the spread is what converts skeptical senior reps — far more than any training deck. Run it on your own data rather than quoting an industry statistic, because your own numbers can't be argued with.

Coaching cadence math. Practical ranges: one recorded discovery call reviewed per rep per week during the first thirty days, dropping to one every two weeks once habits form. Budget 30 to 45 minutes per rep per week for deal-and-skill coaching combined. For a manager with eight reps, that's roughly five hours a week of real coaching — which is the actual reason most coaching programs die. If the manager also carries a number, or has twelve direct reports, the math simply doesn't work and no amount of exhortation fixes it. The honest answer in that case is to change the span of control or accept a lighter cadence, not to pretend.

Ramp effects. New reps are the highest-leverage audience because you're installing a habit rather than replacing one. Building the evidence standard into onboarding — certification on the six elements, three recorded role-plays before they touch a live discovery call — is cheaper than retrofitting it onto a tenured rep who has closed deals without it and reasonably believes they don't need it.

Trade-offs: MEDDIC versus the alternatives, and when to stop

MEDDIC is not universally correct, and coaching it into a team where it doesn't fit produces theater. It was built for complex, high-consideration B2B sales with multiple stakeholders, a real evaluation process, and a budget holder distinct from the user. The further your motion sits from that description, the more the framework costs and the less it returns.

Where it fits poorly. A transactional motion with a two-week cycle, a single decision-maker, and a price point that clears on a credit card doesn't need six elements of qualification — the overhead exceeds the deal value. Product-led motions where the buyer self-serves into a paid tier and sales enters late already have usage data doing the work that Identify pain and Metrics would do. High-velocity SMB teams often do better with something lighter — BANT, or a three-question qualification bar — because the cost of a bad deal is low and the cost of a slow one is high.

Where it fits well and needs extending. For enterprise deals with procurement, security review, and legal, plain MEDDIC leaves out two things that reliably kill deals late: the paper process and competition. That's the case for MEDDPICC. The Paper process element forces the rep to map redlines, security questionnaires, and signature authority *before* the last two weeks of the quarter, which is when those things otherwise surface as surprises. Competition forces an explicit answer to "who else, including the internal build option and doing nothing." Coach the six first; add P and C once reps are consistently earning the original six, because handing a struggling rep eight elements produces worse results than six done well.

Alternatives that solve adjacent problems. Challenger is about *how* you run the conversation — teach, tailor, take control — where MEDDIC is about *what* you must know to believe the deal. They're complementary, not competing; a rep can run a Challenger-style commercial insight and still fail to identify the economic buyer. SPIN is a questioning technique that pairs naturally with the Identify pain and Metrics elements. Command of the Message and similar value-messaging frameworks handle the "why us" that MEDDIC deliberately doesn't address. The Sandler and Miller Heiman lineages overlap on qualification and stakeholder mapping. The trap is running two frameworks in parallel and asking reps to hold both — pick a qualification spine, pick a conversation method, and be explicit about which is which.

The cost side, stated honestly. Coaching MEDDIC properly consumes manager time, adds friction to stage progression, and temporarily shrinks reported pipeline. It also creates a real risk of making discovery feel like an intake interview, which damages the buying experience — the antidote is teaching reps to give before they take: share a benchmark or an observation from a similar account, then ask. And it can be gamed. Any framework with fields can be gamed, which is why the evidence standard, not the field, has to be the unit of enforcement.

When to stop coaching and change something else. If a rep executes MEDDIC flawlessly in rehearsal and refuses to on live calls after two documented follow-up cycles, that's an accountability conversation, not a coaching one. If an entire team resists, look upstream before you look at the reps: broken territories, a comp plan that pays for activity rather than qualified progression, a marketing-sourced pipeline full of leads who were never going to buy, or a sales process whose stages don't match how buyers actually decide. Methodology cannot repair a structurally bad system, and reps know that faster than managers do.

Where this breaks in practice

Treating it as a CRM form. The dominant failure. Reps backfill fields to clear the gate, managers see green dashboards, and everyone is surprised at quarter end. The fix is mechanical: coach the questions, audit the quotes, and make at least one field require a pasted customer statement rather than free text.

Coaching the deal instead of the rep. Saving the Acme deal feels productive and is occasionally necessary, but if you rescue the deal without building the questioning skill, you'll re-coach the identical gap next month on a different logo. Split your 1:1 explicitly: fifteen minutes on the deal, fifteen on the skill the deal exposed.

No follow-through. You agree on two questions for Thursday's call and never listen to Thursday's call. The rep correctly concludes that coaching is a ritual. One skipped follow-up costs more credibility than three good sessions build.

One-size coaching. A skill gap and a will gap need opposite responses — more teaching for one, more accountability for the other. Applying the same warm workshop to both leaves half the team uncoached and the other half resentful.

Logging a friendly as a champion. The most expensive single error, because it inflates forecast confidence on deals with no internal advocate. Apply the intro test every time, and make "champion confirmed" mean "asked for the intro and got it," not "seems supportive."

Exempting top performers. Managers skip MEDDIC coaching for reps who are hitting quota, which teaches the team that the standard is optional and applies only to strugglers. Hold the strongest reps to it too — and then have them run a peer call review, which is the cheapest scaling mechanism available and lands better than manager feedback for many teams.

Confusing pain with a compelling event. A quantified pain with no deadline slips forever. The question "what happens if you do nothing for two more quarters?" exists to surface the event — a contract expiry, an audit, a system sunset, a new executive's mandate. No event, no urgency, no matter how big the number.

Ignoring the RevOps side. If the stage definitions in the CRM don't align with the evidence you're asking for, reps get contradictory instructions from their manager and their system, and the system wins. Ops has to build the gates, the reports that surface evidence gaps, and the win-rate split analysis that proves the whole thing works. Coaching without that scaffolding survives exactly as long as the manager running it.

Skipping the diagnosis and the drill design. Two lightweight practices prevent most of the above. Run a single-element drill each week — "Metrics Week," ten minutes, a fictional buyer, one question that must land on a number with no leading phrasing allowed. And when a deal has no champion, convert it into a live fifteen-minute champion hunt where you play the skeptical stakeholder and the rep practices three moves: naming the gap out loud, asking for access, and testing who really has influence. Both turn abstract feedback into a rehearsed motion, which is the only kind that survives contact with a real buyer.

Related questions

How long before MEDDIC coaching shows up in results?

Leading indicators — metric-capture rate, economic-buyer access — move in two to four weeks. Win-rate impact lags a full sales cycle, so six to twelve months in enterprise. Judge the program on the leading indicators for the first two quarters, not on bookings.

Should MEDDIC fields be required in the CRM?

Require them at stage advance, not at creation. Required-at-creation produces garbage entries. Better: gate Discovery-to-Proposal on three evidence-backed elements — metric, pain with a compelling event, and a tested champion — and sample deals monthly to check the evidence is real.

Who owns MEDDIC coaching, the manager or enablement?

Enablement owns the curriculum, certification, and the shared question bank. The frontline manager owns the weekly deal review, diagnosis, and follow-through. RevOps owns the stage gates and the reporting that makes gaps visible. Programs fail when enablement trains and nobody reinforces.

Does MEDDIC work for renewals and expansion?

Partly. Metrics, economic buyer, and champion translate directly and are often weaker in renewals than in new business, because the original champion left. Decision criteria and process matter less on a routine renewal but matter enormously on a competitive re-evaluation or a multi-year expansion.

Can AI score MEDDIC automatically?

Conversation intelligence tools can flag whether questions were asked and extract candidate answers from transcripts, which scales manager visibility across a whole team. They can't reliably judge whether a champion is real. Use them to find where to look; the diagnosis and the rehearsal stay human.

FAQ

What's the difference between MEDDIC and MEDDPICC?

MEDDPICC adds Paper process — legal, procurement, security review, signature authority — and Competition, including the internal build option and doing nothing. Use MEDDPICC for enterprise deals with formal procurement. Use plain MEDDIC for mid-market and faster cycles. Start with six elements and add the other two only once reps are consistently earning the original six with evidence.

How is a champion different from a coach?

A coach shares information. A champion has organizational power, a personal stake in your success, and sells for you when you're not in the room. The test is access: a real champion introduces you to the economic buyer. If they deflect the request twice, you have a coach — which is useful, but you should stop forecasting the deal as though you have an advocate.

My rep fills in every field but deals still slip. What's wrong?

Almost always the pain or the metric is soft, or there's no compelling event forcing a decision this quarter. Re-run discovery on the pain specifically. A quantified problem with no deadline slips indefinitely regardless of how clean the rest of the record looks. Also check whether the fields contain the rep's language or the buyer's — that distinction usually explains it.

How do I coach this without discovery feeling like an interrogation?

Teach reps to give before they take: share an insight, a benchmark, or a pattern from a similar account, then ask. Spread the six elements across two or three conversations rather than extracting all of them on call one. Pull recordings where the questions landed naturally and have the team study those specific moments instead of a generic script.

What if my managers don't have time to coach weekly?

That's a real constraint, not an excuse. Thirty to forty-five minutes per rep per week is the honest cost; with eight reps that's five hours. If the manager also carries a quota or has twelve directs, reduce the span of control, cut the cadence to biweekly with peer review filling the gap, or use transcript scanning to triage which calls actually need attention. Don't pretend the math works when it doesn't.

Does this apply to teams outside software sales?

Yes, wherever the sale is complex, multi-stakeholder, and budget-approved rather than impulse-bought — industrial equipment, professional services, healthcare systems, construction and facilities contracts. The vocabulary changes and the paper process is often heavier, but the underlying discipline of proving economic value, access to money, and a real internal advocate is identical.

Sources

• MEDDIC Academy — The MEDDIC Sales Methodology
• Force Management — MEDDICC
• Gong — Discovery Call Tips
• RAIN Group — Sales Discovery Questions
• Harvard Business Review — The New Sales Imperative
• Salesforce — Sales Coaching
• Gartner — Sales Practice Insights
• Challenger — The Challenger Sale

Related on PULSE

• [Top 10 MEDDIC Coaching Prompts for Ramping Reps](/knowledge/cg0619)
• [Top 10 MEDDIC Coaching Prompts for Mid-Market Reps](/knowledge/cg0616)
• [Top 10 MEDDIC Coaching Prompts for SMB Reps](/knowledge/cg0617)
• [Top 10 MEDDIC Coaching Prompts for Top Performers](/knowledge/cg0621)
• [Top 10 MEDDIC Coaching Prompts for Underperformers](/knowledge/cg0620)
• [Top 10 MEDDIC Coaching Prompts for Remote Reps](/knowledge/cg0622)