--- name: running-discovery-calls description: >- Plans and executes a B2B discovery call using SPIN question sequencing and Gap Selling diagnosis, producing a pre-call question plan, live-call guardrails, and a post-call current-state/future-state gap write-up with quantified impact. Use when the user says discovery call, disco, first call, qualification call, "what should I ask", discovery questions, call plan, "prep me for this meeting", "review my discovery notes", "I didn't get to pain", root cause, cost of the problem, or pastes a transcript and asks what they missed. Do NOT use for scoring an opportunity against qualification criteria (see qualifying-with-meddpicc), for stalled or fence-sitting buyers late in the cycle (see handling-buyer-indecision), or for close-plan and procurement sequencing (see building-deal-close-plans). metadata: version: "1.0" --- # Running discovery calls Produce the question plan for one upcoming discovery call, and after the call produce a diagnosis of the buyer's current state, future state, and the quantified gap between them. Out of scope: the qualification scorecard, demo scripting, proposal writing, and any output that recommends a product before the gap is quantified. ## The two mechanics you are combining SPIN supplies the question grammar; Gap Selling supplies what the answers must add up to. Keep both terms straight and use them consistently throughout the output. **SPIN** (Neil Rackham, from behavioural analysis of 35,000 sales calls at Huthwaite) sequences four question types ([Sales Enablement Collective](https://www.salesenablementcollective.com/spin-sales-methodology/)): 1. **Situation** — current context and process. "What tool are you currently using for [use case]?" / "Who in the business is responsible for [use case]?" 2. **Problem** — surface pain and test fit. "What's your biggest pain point on a day-to-day basis?" / "What gets in the way of you achieving [important business goal]?" 3. **Implication** — make consequences felt. "How does [pain point] impact your ability to hit KPIs?" / "How much time/money is [issue] costing your company?" / "What goals or successes is [issue] blocking?" 4. **Need-Payoff** — get the buyer to articulate the value of the fix. "What would change if you did [process] differently?" / "What would you do with the time/money [issue] is costing you if it went away?" Implication and Need-Payoff are the pair reps collapse. Implication explores "the negative impacts of the status quo," creating discomfort and urgency, and comes after a problem is identified but before solutions. Need-Payoff highlights "the value of change," creating motivation and buy-in, and only works once urgency exists ([Sales Enablement Collective](https://www.salesenablementcollective.com/spin-sales-methodology/)). Asking Need-Payoff before Implication produces polite agreement and no urgency. **Gap Selling** (Keenan) requires you to map Current State → Future State → the Gap, "the distance between a buyer's current state and their desired future state," where "a wider gap means a stronger motivation to buy" ([Gong](https://www.gong.io/blog/gap-selling)). Current State has five mandatory elements: **Environment, Problem, Impact, Root cause, Emotion** ([Gong](https://www.gong.io/blog/gap-selling)). Future State mirrors them: desired process, impact of solving, desired environment, emotional outcome. Treat discovery as diagnosis, not qualification — the rep moves "from pitcher to problem expert," measures "not just what the problem is, but what it costs them every day it goes unsolved," and only then establishes that the future state outweighs the current one ([A Sales Growth Company](https://salesgrowth.com/gap-selling-book/)). ## Call benchmarks to design against - Ask **11-14 targeted questions** on a discovery call; that range shows the greatest chance of success, and above it success drops to average, from over **519,000 B2B call recordings** ([Gong](https://www.gong.io/blog/nailing-your-sales-discovery-calls)). - **Distribute questions evenly** across the call and do not front-load them; average reps work a checklist at the top of the call ([Gong](https://www.gong.io/blog/nailing-your-sales-discovery-calls)). - Seller talk time was **57% on closed-won deals versus 62% on lost deals** across **326,000 calls** of at least ten minutes; the original 2016 Gong "golden ratio" is **43% talking / 57% listening**, and talking more than **65%** of the call was associated with lower conversion and win rates ([Gong](https://www.gong.io/blog/talk-to-listen-conversion-ratio)). Cite 43:57 as the aspirational target and 57% as the observed won-deal average; they are different studies. - More questions is not better: sellers who **won asked 15-16 questions**, sellers who **lost asked about 20** ([Gong](https://www.gong.io/blog/talk-to-listen-conversion-ratio)). Over-questioning reads as interrogation. ## Where the call sits in the sequence SPIN defines four call stages, and only the first two belong to discovery ([Sales Enablement Collective](https://www.salesenablementcollective.com/spin-sales-methodology/)): | Stage | What happens | Rule | |---|---|---| | Opening | Rapport plus Situation questions | Not a pitch; no product mention | | Investigating | Problem questions; qualify fit | The bulk of the call belongs here | | Demonstrating Capability | Benefits over features, driven by Implication and Need-Payoff | Only after the gap is quantified | | Obtaining Commitment | Stakeholders, buying process, objections | A separate meeting on complex deals | There is no fixed timeline: a rep may cover two stages in one call, or need weeks to reach the last ([Sales Enablement Collective](https://www.salesenablementcollective.com/spin-sales-methodology/)). Decide which stages this call covers before drafting questions, because a plan that tries to reach Commitment in a first call spends its question budget on process rather than diagnosis. ## Workflow Copy this checklist into your reply and tick items as you complete them: ``` - [ ] 1. Establish the call's single diagnostic objective - [ ] 2. Draft 11-14 questions mapped S / P / I / N and spread across the call - [ ] 3. Run the plan audit (coverage + count + sequence); fix and re-run until clean - [ ] 4. Add the live-call guardrails block - [ ] 5. After the call: fill the five Current State elements from the transcript - [ ] 6. Quantify the gap in the buyer's own units - [ ] 7. Emit the diagnosis in the output template ``` **1. Set one diagnostic objective.** Write it as a hypothesis to be disproved, for example "I believe renewals slip because entitlement data lives in three systems." A call with more than one objective produces shallow answers on both, because Implication questions need three or four turns on a single thread to land. **2. Draft the questions.** Write 11-14 and label each S, P, I, or N. Use Gap Selling's **Process** questions to find root cause — "Can you help me understand how your company currently does [process]?" / "Can you walk me through what happens before/after [process]?" — and **Validating** questions to confirm — "To me, it seems like [problem] is the main issue here, am I understanding you correctly?" ([Gong](https://www.gong.io/blog/gap-selling)). Adapt the wording freely to the buyer's vocabulary; the sequence is what is load-bearing, not the phrasing. **3. Plan audit — validate, fix, re-validate.** Check the draft against all four tests: (a) total is 11-14; (b) at least two Implication questions and at least one Need-Payoff; (c) no Need-Payoff appears before an Implication; (d) each of the five Current State elements has at least one question that can reach it. Fix any failing test and re-run all four, because fixes usually break the count. Only proceed when all four pass. **4. Add guardrails.** Append the live-call block: target seller talk time at or below the 57% won-deal average; pause two seconds before responding so the buyer elaborates; paraphrase and confirm ("So what I'm hearing is that efficiency is your top concern. Does that sound right?"); use "tell me more" as the default follow-up ([Gong](https://www.gong.io/blog/talk-to-listen-conversion-ratio)). **5. Post-call: fill the five elements.** From the transcript or notes, fill Environment, Problem, Impact, Root cause, and Emotion. Mark any element you cannot fill with a buyer quote as `UNKNOWN` and write the exact question to ask next time. Never infer Root cause from Problem; the two are routinely different, and the whole diagnosis fails if you fix a symptom. **6. Quantify.** Convert Impact into a number using the buyer's own operating figures, not industry averages. Worked pattern: 20 reps × 5 hrs/week manual entry = 100 hrs/week = 5,200 hrs/year × $50/hr fully loaded = **$260,000/year** ([Weflow](https://www.weflow.ai/blog/neat-selling)). Show the arithmetic so the buyer can dispute an input rather than reject the total. If you lack an input, state the missing input and stop; do not substitute a benchmark, because a number the buyer cannot recognise ends the credibility of the whole business case. **7. Emit the diagnosis** using the template below. ## Output format Use this exact section order for both outputs. Question wording and cell contents are yours to adapt; the labels and the order are fixed so that call notes stay comparable across a team. ```markdown ## Pre-call plan — / / Objective (hypothesis to disprove): | # | Type | Question | Reaches | |---|---|---|---| | 1 | S | What tool are you currently using for ? | Environment | | 2 | P | What gets in the way of ? | Problem | | 3 | I | How much time is costing the team each week? | Impact | | ... | | | | Guardrails: talk <=57% · two-second pause · paraphrase and confirm · "tell me more" ## Post-call diagnosis — / **Current State** - Environment: - Problem: <...> - Impact: <...> - Root cause: <...> - Emotion: <...> **Future State** - Desired process / impact of solving / desired environment / emotional outcome **The Gap** **Next question set** ``` ## Gotchas - Do not open with a reframe or point of view. A Challenger-style reframe delivered before you understand the situation "is a guess," and the technique "works best in the second half of a discovery conversation, not the first 5 minutes" ([PitchMonster](https://www.pitchmonster.io/blog/challenger-sale-methodology)). Earn standing with Situation and Problem questions first. - Talk ratio is a consistency signal, not just a volume one. High performers hold roughly the same ratio whether they win or lose; low performers swing ten points, from 54% talk time on won deals to 64% on lost ones ([Gong](https://www.gong.io/blog/talk-to-listen-conversion-ratio)). If your post-call review shows a big swing between good and bad calls, the fix is process discipline, not a lower target. - Solving the stated need is a failure mode, not a win. "Never sell to need. If you only solve the problem your buyer thinks they have instead of the one they really have, you haven't helped them at all" ([A Sales Growth Company](https://salesgrowth.com/gap-selling-book/)). Push two or three layers past the first answer with repeated "why" before accepting a Problem. - Separate symptom from cause explicitly in the notes. "We can't trust our pipeline" is a symptom; "activity data is incomplete because reps don't log meetings" is a cause ([Weflow](https://www.weflow.ai/blog/neat-selling)). Writing the symptom in the Root cause field is the single most common defect in discovery records. - Emotion is a required Current State element, not soft colour ([Gong](https://www.gong.io/blog/gap-selling)). Capture who is personally exposed by the problem and how, because "all sales are about change... every sale is emotional" ([A Sales Growth Company](https://salesgrowth.com/gap-selling-book/)) and the exposed person is your candidate champion. - Validate the pain with more than one stakeholder before treating it as established ([Weflow](https://www.weflow.ai/blog/neat-selling)). A single contact's account of the process is their view of it, and root causes usually sit in a team they do not run. - SPIN neither qualifies the deal nor supplies a point of view ([PitchMonster](https://www.pitchmonster.io/blog/challenger-sale-methodology)). When the user's real question is whether the deal is real, stop and route to qualifying-with-meddpicc rather than adding more questions. - Do not front-load the question block even when the buyer is cooperative, and do not read the plan as a script. Working a checklist at the top of the call is the signature of an average rep in the Gong sample ([Gong](https://www.gong.io/blog/nailing-your-sales-discovery-calls)); the plan is a coverage map, and the order should follow whatever thread the buyer opens.