What are the 10 concrete steps from *SPIN Selling* for executing a discovery call in 2027?

The call that goes nowhere, and the one that does

Two reps at the same RevOps software company run discovery on the same account in the same week. Both get forty-five minutes with a VP of Revenue Operations at a 900-person logistics firm. Both calls are recorded, both get transcribed, both land in the CRM with a "Discovery Complete" stage change. Only one of them turns into anything.

The first rep opens with a company overview slide, then spends nineteen minutes asking questions the buyer has already answered in a form: how many reps do you have, what CRM are you on, who owns the data, what's your fiscal year. The buyer answers each one politely and gets progressively flatter. At minute thirty the rep pivots to a demo of the forecasting module. At minute forty-two the buyer says "this is interesting, send me some materials, let me talk to my team." The rep logs it as a strong call. Nothing happens. Three follow-up emails go unanswered.

The second rep spends four minutes on background she could not have looked up — how the last forecast cycle actually went, who ended up rewriting the numbers, what happened when the CFO asked why. Then she asks what breaks when the pipeline data is wrong. The buyer says reps update opportunities the night before the forecast call. She asks what that means for the number the VP presents. He says he pads it. She asks what happens downstream when he pads it — how hiring plans, quota setting, and board commitments get built on a padded number. He goes quiet, then talks for six straight minutes about a headcount freeze that traced back to a blown Q2 forecast. She asks what it would be worth to walk into the forecast call knowing the number was real. He tells her. Then he asks who else at his company needs to see this, and books a working session with his RevOps manager and the CFO's analyst for the following Tuesday.

Nothing about the second call was magic. It was the SPIN sequence, executed in order, with the discipline to not skip from Problem straight to product. Neil Rackham's research at Huthwaite, published in *SPIN Selling* in 1988, found that this pattern separated high performers from average ones in complex, multi-call sales — and that the behaviors which worked in small transactional sales actively hurt in large ones. Nearly four decades later the pattern has held up better than almost any other sales framework, because it describes how buyers think rather than how sellers want to talk.

What follows are the ten concrete steps for executing a discovery call on that model, adapted for how buying actually works in 2027 — with AI-assembled account research, buyers who have already read your docs, and committees of eight to eleven people who will never all be on one call.

How the mechanism actually works

The sequence matters more than any individual step. SPIN is not a menu of question types you sprinkle across a conversation. It is an escalation: each stage only works because the previous one built the ground it stands on. Skip a stage and the next one lands as an interrogation or a pitch.

Step one — do the research so you can skip Situation questions. Before the call, assemble what is publicly knowable: headcount by function, recent leadership changes, tech stack signals, funding or earnings language, org structure, hiring posts (a job req for a "Revenue Operations Analyst — forecasting" is a stated problem in disguise), and any public commentary from the person you're meeting. In 2027 this is the one step that has genuinely changed. Rackham wrote in an era where you called to find out how many salespeople the company had. That fact is now assembled automatically in the minutes before the call. The strategic consequence is that every Situation question you ask which could have been researched now costs you credibility instead of buying you time. Budget fifteen to thirty minutes of prep for a first discovery call on a mid-market account, more for enterprise. Write down three hypotheses about what is probably broken. You will test them, not assert them.

Step two — open by establishing purpose and earning the right to ask. Rackham's Preliminaries stage is short and its only job is to get you permission to ask questions. Two things belong in it: why you're here in one or two sentences, and an explicit ask for the floor. "My understanding is you're rebuilding how forecasting works ahead of next fiscal year. I'd like to spend most of our time understanding how it works today rather than showing you slides — is that a fair use of the next forty minutes?" Do not open with a company overview. Rackham's data was blunt on this: opening with capability statements correlates with worse outcomes in large sales because it invites objections before you know what the buyer values. Two to four minutes. Then stop talking.

Step three — ask the minimum viable Situation questions. Situation questions gather background facts. They are necessary and they are dangerous: Rackham found that high performers ask *fewer* of them than average performers, because each one costs the buyer effort and returns the buyer nothing. Ask only what you could not research and what you actually need to interpret the answers that follow. Good ones are about process and sequence rather than facts: "Walk me through what happens between a rep updating an opportunity and a number appearing on your forecast." Bad ones are database lookups: "How many reps do you have?" Aim for three to five Situation questions in a forty-five minute call. If you're at ten, you did not do step one.

Step four — surface Problem areas. Problem questions ask about difficulties, dissatisfactions, and things that don't work. "Where does the forecast usually go wrong?" "What part of this process do people complain about?" "How reliable is the data by the time it reaches you?" These uncover what Rackham called Implied Needs — statements of a problem or dissatisfaction. In small sales, uncovering implied needs is enough to close. In large sales it is not, and this is the single most common place where discovery calls stall out: the rep hears a problem, gets excited, and starts solving. Do not solve. Collect two or three real problems and pick the one with the most downstream weight.

Step five — develop Implications until the problem gets expensive. This is the highest-skill move in the entire framework and the one Rackham's research identified as most strongly associated with success in large sales. Implication questions take a problem the buyer has admitted and ask what it causes: what it costs, who else it affects, what it prevents, what happens if it continues. "You said reps update the night before — what does that do to the accuracy of the number you present?" "When the number's off, what happens to hiring plans?" "How does that affect your credibility with the CFO?" Each question makes the problem bigger and more urgent in the buyer's own mind, using the buyer's own words. Implication questions are uncomfortable to ask because they make the buyer feel worse, which is exactly why average performers avoid them. Plan three to five in advance, written down, tied to your pre-call hypotheses.

Step six — build Need-payoff so the buyer states the value. Need-payoff questions flip the emotional direction: they ask about the usefulness, value, or benefit of solving the problem. "What would it be worth to walk into that forecast call confident in the number?" "If your reps' pipeline data were clean by Thursday, what would that free you up to do?" "Who else would care if this got fixed?" The point is not to make the buyer feel good. The point is to get the buyer to articulate the benefits out loud, in their language, so that when they later repeat it to a CFO or a committee, they are repeating their own argument rather than your pitch. Rackham found these questions also reduce objections, because a benefit the buyer stated is one they will not argue against.

Step seven — confirm the Explicit Need in the buyer's words. Before you show anything, play it back and get agreement: "So if I've got this right — the core issue is data hygiene at the rep level, and the cost of it is a forecast you have to pad, which lands on headcount planning. And what you'd want is a number you trust by Thursday of week one. Is that fair?" This converts an implied need into an Explicit Need — a clear statement of want or desire — and Rackham's data showed Explicit Needs are the strongest predictor of purchase in large sales. It also gives you a documented, buyer-approved problem statement to reference in every subsequent touch.

Step eight — demonstrate capability as Benefits, not Features or Advantages. Rackham drew a hard line here. A Feature is a fact about your product. An Advantage is how the feature could be useful in general. A Benefit shows how the feature meets an Explicit Need the buyer has already stated. In large sales, Features are neutral-to-negative, Advantages are neutral, and Benefits are what correlates with success. So do not demo the whole platform. Show one or two things that map directly to the need you just confirmed, and narrate the mapping explicitly: "You said you need the number solid by Thursday — here's the piece that forces field completion before an opportunity can advance." Ten minutes, maximum, on a discovery call. If the buyer asks for more, that's a next meeting.

Step nine — obtain an Advance, not a Continuation. Rackham's four outcomes are Order, Advance, Continuation, and No-Sale. An Advance is a specific action that moves the sale forward on a date. A Continuation is a call where nothing was agreed but everyone was pleasant — "send me materials," "let's touch base next month." Continuations feel like success and are functionally No-Sales. Plan your intended Advance before the call and have a smaller fallback ready. Good advances: a working session with a named second stakeholder on a specific date; access to a sandbox with a specific dataset; a scoping call with the technical owner; a written problem statement you draft and they edit. Ask for it directly before the clock runs out. Leave five minutes.

Step ten — debrief while it's fresh and update the account strategy. Within thirty minutes, write down: the problems in the buyer's exact words, the implications they confirmed, the need statement they agreed to, the advance and its date, who else got named, and what you now believe about the deal that you didn't before. This is where the modern tooling earns its keep — a recorded, transcribed call gives you verbatim language for the follow-up email and for the internal brief that arms your champion. Rackham's own recommendation was to practice one SPIN behavior at a time and review after each call; that discipline is what turns the framework from a script into a skill.

Real numbers, ranges, and what to actually measure

The framework is qualitative but execution is measurable, and the measurements are what let you coach it.

Talk-time ratio. Conversation-intelligence tooling — Gong, Chorus, Clari Copilot and their peers — has made rep talk-time a standard reported metric. The practical target on a discovery call is a rep talking substantially less than half the time; roughly 30–45% is a reasonable working band, with the buyer holding the rest. The number matters less than the direction of travel: if your talk-time on discovery calls is above 60%, you are pitching, not discovering, and no amount of question-list training will fix it until that inverts. Pull your last ten discovery recordings and get the actual number before you change anything.

Question mix by type. Tag your questions across a call and count them. A workable distribution for a forty-five minute first discovery on a mid-market deal: 3–5 Situation, 5–8 Problem, 3–5 Implication, 2–4 Need-payoff. The Implication count is the diagnostic. Most reps who think they're running SPIN are actually running Situation-Problem-Demo, with zero Implication questions in the transcript. If a coach can't find three Implication questions in a recorded call, the call did not run SPIN regardless of what the CRM stage says.

Longest monologue. Two numbers are worth tracking: your longest uninterrupted stretch of talking, and the buyer's. On a good discovery call the buyer's longest monologue should exceed yours, often by a wide margin. The six-minute buyer monologue in the opening scenario is the signal you're looking for — it almost always arrives immediately after a well-placed Implication question.

Time allocation. A rough forty-five minute skeleton: 2–4 minutes preliminaries, 5–8 minutes Situation, 10–15 minutes Problem and Implication together (this is the heart of the call), 5–8 minutes Need-payoff and confirmation, up to 10 minutes capability demonstration, 5 minutes advance and next steps. Notice that the demo is the smallest block and the last content block. Most reps invert this.

Multi-threading. Enterprise buying groups have grown steadily for a decade; large B2B purchases routinely involve six to ten-plus stakeholders. One consequence for discovery: a single call cannot possibly surface the full need set, so a named second stakeholder is the single most valuable Advance you can get. Track "stakeholders named per discovery call" and "stakeholders met by day 30" as leading indicators. A deal with one contact at day thirty is a deal in trouble, and no amount of good questioning of that one contact changes it.

Prep time. Fifteen to thirty minutes for mid-market, forty-five to ninety for enterprise. This has not shrunk as much as people expect in the AI era. What the tooling removed was the *gathering* — the reading, the tab-hopping, the LinkedIn archaeology. What it did not remove was the *thinking*: forming three hypotheses about what's broken and drafting the Implication questions that test each one. That thinking is the prep. If your AI briefing tool gave you a summary and you spent zero minutes forming hypotheses from it, you have automated the wrong half.

Advance rate. The cleanest single metric for discovery quality: what percentage of discovery calls end with a specific, dated, agreed next action with a named participant. Not "they were interested." Not "sent the deck." A date and a name. Whatever your current number is, it is almost certainly lower than your CRM suggests, because Continuations get logged as progress. Audit thirty closed-lost deals and count how many had a real Advance out of first discovery.

Coaching cadence. Rackham's guidance was to practice a single behavior at a time, try it in safe accounts first, and focus on quantity of attempts before quality. Translated to a modern team: one behavior per rep per two-week cycle, one recorded call reviewed per week against that single behavior, with the reviewer counting instances rather than giving impressions. "You asked one Implication question" is coachable. "The call felt a bit pitchy" is not.

Trade-offs, adjacent frameworks, and when SPIN is the wrong tool

SPIN is a questioning methodology for complex sales. It is not a qualification framework, not a deal-management framework, and not a negotiation framework, and most of the friction teams experience with it comes from asking it to be one of those.

SPIN versus qualification frameworks. BANT (Budget, Authority, Need, Timeline), MEDDIC/MEDDPICC, and their relatives answer "should we pursue this deal and what do we still not know?" SPIN answers "how do I run the conversation?" They are complements, not substitutes, and the common failure is running a qualification checklist *as* the discovery call. A rep working down a MEDDIC field list out loud produces exactly the flat, interrogative call from the opening scenario. Run SPIN in the conversation; fill the qualification fields from what the conversation produced. If a MEDDIC field is still blank after the call, that's a target for the next one, not a question to force into this one.

SPIN versus Challenger-style teaching. The Challenger model argues for leading with commercial insight — teaching the buyer something about their business, tailoring it, and taking control of the conversation. SPIN's Implication and Need-payoff questions do related work but arrive from the opposite direction: rather than telling the buyer what their problem costs, you ask until they say it. The trade-off is real. Teaching is faster and works well when the buyer genuinely doesn't know what they don't know, or when you're creating a category. Asking is slower but produces a buyer-owned argument that survives the internal committee meeting you won't be in. Many effective reps blend them: teach to establish the problem exists in their industry, then use Implication questions to make it specific and local. What does not work is teaching in place of asking, because you end up with a buyer who agrees with your framing and cannot reproduce it.

SPIN in small or transactional sales. Rackham was explicit that his findings applied to large, multi-call sales, and that behaviors effective in small sales — including strong opening benefit statements and aggressive closing techniques — were neutral or counterproductive in large ones. The reverse also holds. If you sell a $40/month product with a two-call cycle, a full Implication sequence is overkill and will annoy people. Use a compressed version: one Problem question, one Implication, straight to value. Match the depth of questioning to the size of the decision.

SPIN when the buyer arrives already educated. This is the 2027-specific trade-off. Buyers routinely complete most of their research before contacting a vendor; many arrive having read your documentation, your pricing page, competitor comparisons, and AI-generated summaries of all three. Running a full Situation sequence on that buyer is insulting. The adaptation is to compress steps three and four hard and start close to Implication: "You've clearly looked at this closely. Rather than covering ground you've already covered — what made you start looking?" Then go straight for what the problem is costing. The framework survives; the front end shortens.

SPIN in product-led and self-serve motions. When usage data precedes the conversation, you can enter discovery with observed behavior instead of hypotheses: which seats activated, which feature stalled, where the workflow drops off. That's a stronger starting point than any research step, and it changes the Problem question from "what's not working?" to "I noticed twelve people activated and four are still active in week three — what happened to the other eight?" Same sequence, better inputs.

The cost of doing it right. Well-run SPIN discovery is slower per call and per deal in the early stages. You will book fewer demos out of first calls. Pipeline created will look worse for a quarter while the pipeline that does get created converts better. If your comp plan and your management dashboards reward demos-booked or opportunities-created, your reps will rationally skip Implication questions and pitch early, regardless of what the training said. Fixing discovery is partly a training problem and largely an incentive problem — the strategy has to change at the dashboard level or the behavior won't.

Where discovery calls actually break

Solving at the first problem. The buyer says something painful, the rep's product solves it, and the rep jumps. This kills more discovery calls than any other single behavior. The problem is that an Implied Need — "our forecast is unreliable" — is not enough to justify a purchase in a large sale. By solving immediately you trade a potentially large, expensive, board-visible problem for a small feature conversation. The fix is mechanical: when you feel the urge to solve, ask one more Implication question instead. Write "ONE MORE" on a sticky note if you have to.

Situation-question bloat. Reps ask background questions because they're safe — nobody gets uncomfortable, the call feels productive, and the CRM fields get filled. Rackham's finding that top performers ask fewer of them is counterintuitive precisely because they feel useful. The 2027 version of this failure is worse: the information was in the AI briefing that the rep didn't read. Rule of thumb — if the answer is on the company's website, in their job postings, or in your enrichment data, asking it costs you credibility for zero information.

Implication questions that sound like an attack. Done clumsily, "so what does that cost you?" repeated four times feels like being cross-examined. Three defenses: use the buyer's own words back to them, alternate implication with genuine acknowledgment, and preface with a third-party frame when the topic is sensitive ("something we hear from other RevOps leaders is that padding the forecast eventually lands on hiring plans — does that track here?"). Tone carries this stage entirely.

Confusing a Continuation for an Advance. Covered above but it deserves repeating because it distorts every downstream number. "They loved it, sending the deck" is not progress. Build the distinction into your CRM: require a date and a named participant to move a deal past discovery, and audit stage changes that lack them.

Single-threading. A great discovery call with one person in an eight-person buying committee is a good conversation and a fragile deal. Make "who else is affected by this?" a standing Implication question — it does double duty, deepening the problem and mapping the committee at the same time.

Running the sequence as a script. SPIN describes a pattern of behavior, not a fixed order of scripted lines. Buyers jump around. You will ask a Problem question and get an implication for free; you will need to loop back to Situation mid-call when something surprising surfaces. What must hold is the escalation logic — you don't demonstrate capability before you have a confirmed Explicit Need. Everything else flexes.

Letting the tooling do the thinking. AI briefings, auto-generated question lists, and real-time call assist are genuinely useful for the gathering and the recall. They are not useful for judgment about which of three problems has the most downstream weight, or for reading that a buyer just went quiet because you touched something real. Reps who outsource the hypothesis-forming step arrive with generic questions and generic questions get generic answers.

No debrief. The recording exists; nobody watches it. The single highest-ROI change most teams can make is a fifteen-minute weekly review of one recorded discovery call per rep, counting one specific behavior. Not general feedback — counting. Behavior change follows measurement.

Related questions

Does SPIN Selling still work with AI-assisted buyers?

Yes, with a shortened front end. Research and Situation questions compress because buyers arrive educated and reps get automated briefings. Implication and Need-payoff — the parts that make a problem feel expensive and let the buyer articulate value — are unchanged, because they operate on the buyer's judgment, not their information.

How is SPIN different from MEDDIC?

SPIN is a conversation methodology: what to ask, in what order, during a call. MEDDIC is a qualification and deal-inspection framework: what you must know to forecast the deal. Run SPIN in the room and populate MEDDIC fields from what it produced. Reading MEDDIC fields aloud as questions produces a flat, interrogative call.

How many questions should a discovery call include?

Roughly 13–22 in forty-five minutes, weighted toward Problem and Implication: 3–5 Situation, 5–8 Problem, 3–5 Implication, 2–4 Need-payoff. The count matters less than the mix. Zero Implication questions in a transcript means the call was not SPIN, whatever the CRM stage says.

What counts as a good Advance out of first discovery?

A specific, dated action with a named participant: a working session with a second stakeholder next Tuesday, sandbox access with their data, a scoping call with the technical owner. "Send materials" and "touch base next month" are Continuations — they feel like progress and are functionally dead ends.

Should reps use a written SPIN question list on calls?

Yes for preparation, sparingly during the call. Write three to five Implication questions tied to your pre-call hypotheses. Reading them in order produces an interrogation; having thought them through produces the ability to ask the right one when the buyer opens the door.

FAQ

**Who wrote *SPIN Selling* and what was the research behind it?** Neil Rackham, published in 1988, based on research conducted through Huthwaite that observed and analyzed a large number of live sales calls across many countries. The central finding was that behaviors correlating with success in small, single-call sales were different from — and sometimes counterproductive in — large, multi-call sales, and that questioning patterns rather than closing techniques distinguished top performers.

What do the four letters in SPIN stand for? Situation, Problem, Implication, Need-payoff. They describe four types of question, and the sequence is the point: Situation gathers background, Problem surfaces difficulties, Implication develops what those difficulties cost, and Need-payoff gets the buyer to articulate the value of solving them. Implication is the highest-skill and most commonly skipped stage.

Is a discovery call the same as a demo? No, and merging them is the most common structural mistake. A discovery call's purpose is to surface and develop needs and secure an Advance. Capability demonstration belongs at the end and only against a confirmed Explicit Need — ten minutes maximum on a first call. If the buyer wants a full demo, that is your Advance, not your agenda.

What is the difference between an Implied Need and an Explicit Need? An Implied Need is a statement of a problem or dissatisfaction — "our forecast is unreliable." An Explicit Need is a clear statement of want or desire — "we need a forecast I can trust by Thursday." Rackham's research found Explicit Needs strongly predict purchase in large sales; Implied Needs do not, which is why moving between them via Implication questions is the core skill.

How should the ten steps change for a $5,000 deal versus a $500,000 one? Compress proportionally. Small deals: one Problem question, one Implication, straight to value, close in the call. Large deals: full sequence, multiple stakeholders, and the expectation that no single call surfaces the whole need set — the strategy becomes a series of discovery conversations, each securing an Advance to the next person in the buying group.

What is the fastest way to tell if a rep is actually running SPIN? Pull a recorded call and count Implication questions. If there are fewer than three, the rep ran Situation-Problem-Demo. Then check the ending: if the agreed next step lacks a date and a named person, they took a Continuation. Those two checks catch most of it in under ten minutes per call.

Sources

• https://www.huthwaiteinternational.com/
• https://en.wikipedia.org/wiki/SPIN_selling
• https://hbr.org/2012/07/the-end-of-solution-sales
• https://www.gartner.com/en/sales/insights/b2b-buying-journey
• https://www.gong.io/resources/
• https://hbr.org/2017/03/the-new-sales-imperative
• https://www.mheducation.com/
• https://www.salesforce.com/resources/research-reports/state-of-sales/

Related on PULSE

• How to structure a 45-minute discovery call agenda
• MEDDPICC vs BANT: which qualification framework fits your deal size
• What separates an Advance from a Continuation in pipeline stages
• How to use conversation intelligence data to coach discovery skills
• Multi-threading enterprise deals: mapping a buying committee from one contact
• Building pre-call research briefs that produce hypotheses, not summaries