Sales

Free discovery call question builder for B2B sales

Name: Discovery Call Question Builder
Author: Charles Summers

By Charles Summers · Updated 8 August 2026 · Free, no signup

Short answer

This builds a sequenced discovery call rather than a list of questions. Pick a methodology and the structure genuinely changes: SPIN runs situation to problem to implication to need-payoff, MEDDIC weights the call towards decision process and metrics, Challenger puts a reframe before any question at all. Deal size sets call length, question count and whether procurement, security and legal appear. Every question is tagged as gathering information or creating realisation, the mix is counted, and minutes per stage plus your airtime budget are computed from the call length.

Use the discovery call question builder

What does this tool actually do?

It runs entirely in your browser. Nothing you type is sent to a server, no account is required, and there is no usage limit, because there is no cost per run to control.

Two kinds of question, and only one of them changes anything

A gathering question moves information from the buyer to you. A realisation question moves information from the buyer to the buyer. Both are necessary and they are not interchangeable, but a call made entirely of the first kind is an interview, and the buyer leaves it having learned nothing they did not know when they answered the phone. Nobody buys because a vendor now understands them well. They buy because a conversation made a cost visible that had previously been absorbed quietly.

The mechanical difference is what the answer does. "How many people are in the support team?" produces a fact you write down. "When a ticket gets reassigned three times, who is chasing it, and what does the customer see while that is happening?" produces a pause. The second question contains no information about your product and no claim of any kind, yet it forces the buyer to assemble a picture that was previously stored in six separate places, one per person who deals with a fragment of it.

Realisation questions fail if they run out of order, which is the most common structural error in discovery. Asking what the impact is before the buyer has said out loud that a problem exists produces a shrug, because you have asked them to price something they have not agreed they own. The sequence is not decoration: state, then problem, then consequence, then their own statement of what solving it is worth. Skip a rung and the ladder does not hold weight.

There is also a hard rule about gathering questions worth applying before the call. Anything answerable from their website, their careers page, their annual report or their product changelog is not a discovery question, it is a demonstration that you did not prepare, and it is charged against a patience budget that is smaller than most reps assume. Convert those into confirmations instead: say what you found and ask what it is missing. That takes fifteen seconds, signals preparation, and usually produces a correction that is more interesting than the original question would have been.

The three methodologies answer three different questions

They are routinely treated as interchangeable brands of the same product, which is why teams end up running all three badly at once. MEDDIC is a qualification framework: it exists to tell you whether this deal deserves more of your quarter, and its output is a scored gap list, not a conversation. SPIN is an interview structure: it exists to make a problem feel expensive enough to act on, and its output is the buyer saying the payoff sentence themselves. Challenger is a teaching sequence: it exists to change what the buyer believes the problem is, and its output is a reframe they had not previously considered.

That difference determines where the framework sits in the call. SPIN is the call. MEDDIC is what you fill in afterwards from what the call produced, which is why running it as an agenda goes so badly: asking who signs the contract at minute four is a question you have not yet earned, and the answer will be politely wrong. Challenger inverts the order entirely, putting the insight before the questions, because its whole claim is that a buyer who has already framed their own problem will only ever ask you to solve it their way.

The consequence for question design is concrete. Under SPIN, most of the airtime belongs to implication questions, and situation questions should be pushed almost entirely into pre-call research. Under MEDDIC, the questions that matter most are the ones about the buying process, which is precisely the section reps skip because it feels commercially forward. Under Challenger, the first two minutes contain a claim rather than a question, and everything after it is testing whether the claim landed. Running the same fifteen questions and relabelling the sections achieves nothing at all.

MEDDIC produces a gap list. Metrics, economic buyer, decision criteria, decision process, identified pain, champion. Each one is either evidenced or it is not, and the deal is forecast on the count.
SPIN produces a sentence from the buyer. If they have not said what solving it would be worth, the need-payoff stage did not happen, whatever questions got asked.
Challenger produces a changed frame. The test is whether the buyer repeats your reframe back to a colleague later in their own words, which is the only version of it that will survive the internal meeting.
No methodology still needs a spine. Problem, quantified consequence, dated event, named process, agreed next step. Five facts. A call that produced fewer than four of them was a pleasant conversation.

Talk ratio is a symptom, and coaching it directly makes calls worse

Conversation analytics vendors publishing from very large transcript sets consistently find that on discovery calls the more successful reps take under half the talk time, usually somewhere in the mid-forties by percentage. The figure is real, the causation is backwards, and the coaching that follows from it is usually harmful. A rep told to talk less produces a call with gaps in it, not a call with better questions. Talk share falls on its own when the questions improve, because a good question buys ninety seconds of answer and a weak one buys eight.

Three secondary numbers are more useful because they are directly fixable. The first is your longest single monologue. Somewhere around a minute and a half of uninterrupted vendor speech, attention starts leaking, and unlike talk share this is a specific moment you can find in a recording and cut. The second is question density in the first ten minutes, which is where a call is either framed as a conversation or as a presentation; if the first ten minutes contain two questions, the remaining thirty will not recover. The third is what happens after their answer stops: the two seconds of silence you leave before speaking is the cheapest tool available and the one reps abandon first under pressure.

Treat the published ranges as calibration rather than targets, and expect them to move with deal size and seniority. A twenty-minute transactional call with an owner-operator legitimately runs closer to an even split, because they want to know what it does before they will spend attention on their own situation. An hour with a committee at a large enterprise runs much further towards them. Judging both against a single number tells you nothing about either.

Compelling events and buying processes are the two things reps do not ask about

A compelling event needs three parts and most logged events have only one. There has to be a date, an owner, and a consequence that lands on someone specific if the date passes. "They want to improve reporting this year" has none. "The current contract terminates on the last day of March and the finance team cannot close the quarter without a replacement in place" has all three, and it tells you when to apply pressure and when pressure is simply noise. The quickest test is to ask what happens the day after the date. If the answer is nothing much, you have a preference with a deadline attached to it.

Buying process questions get skipped because they feel presumptuous, and skipping them is why forecast dates slip in the last two weeks rather than the first six. The reliable version is historical rather than hypothetical. Do not ask what the process is; ask what happened the last time they bought something of this size. People describe the official process and then follow the actual one, and the actual one contains the two-week wait for the monthly finance meeting, the security review nobody mentioned, and the fact that anything above a certain figure goes to a director who has not been in any of your meetings.

Then ask about the signature threshold explicitly, because it is a fact rather than an opinion and it reorganises the deal instantly. A proposal that lands one pound above an approval limit acquires an extra approver and several weeks, and you would rather discover that while the number is still adjustable. The last question of the call is the same regardless of methodology: what has to happen between now and a decision, in order, with names and dates against each step. A buyer who cannot answer that is not obstructing you, they usually have not thought about it, and building it together is the most useful thing you can do in the final five minutes.

Numbers worth knowing

Metric	Typical	What it means
Rep talk share on successful discovery calls	usually under half, often mid-40s%	From conversation analytics vendors' transcript sets. Correlation, not instruction: talk share drops because the questions got better, not because the rep decided to speak less.
Questions asked on a discovery call	commonly reported sweet spot 11 to 14	One widely cited vendor analysis. Above the range calls read as interrogation, below it as presentation. Scale it with call length rather than treating it as a fixed quota.
Longest uninterrupted vendor monologue	keep under about 90 seconds	A coaching threshold rather than a measured constant. Unlike talk share it is a specific timestamp in a recording, which makes it the more useful thing to review.
Compelling events that survive the three-part test	a minority of those logged	Date, owner, consequence. Most CRM entries have a date and an aspiration. Ask what happens the day after the date and the weak ones fail immediately.
Discovery questions answerable from public sources	target zero	Turn each one into a confirmation instead. It costs fifteen seconds, proves preparation, and the correction you get back is usually more useful than the answer would have been.

Mistakes that quietly cost you results

Asking about impact before the buyer has admitted the problem: You are asking them to price something they have not agreed they own, so the answer is a shrug and the question is spent. Establish the problem in their words first, then ask what it costs, in that order, every time.
Running MEDDIC as the agenda for the call: MEDDIC is a qualification scorecard filled in after a conversation, not a script for one. Asking who signs at minute four gets a polite and inaccurate answer. Earn the process questions, then ask them all properly in the last third.
Asking what their buying process is: You get the official version. Ask what actually happened the last time they bought something this size, which surfaces the finance meeting that only sits monthly, the security review nobody mentioned, and the approval threshold that adds an approver you have never met.
Logging a compelling event that is really a preference: Test it: date, owner, and a consequence that lands on a named person. If nothing happens the day after the date passes, you have a wish with a deadline, and pressure applied against it reads as manufactured because it is.
Coaching a rep to reduce their talk ratio: You get pauses, not discovery. Fix the input instead: cut the longest monologue, raise question density in the first ten minutes, and leave two seconds after each answer. Talk share moves on its own once those change.

What does the output look like?

This is the exact output the tool produces from the example inputs. It is generated by the same code that runs when you click the button, so what you see here is what you get.

DISCOVERY CALL PLAN Selling: a claims automation platform for regional insurers Buyer: Director of Claims Operations | Deal band: $25k to $100k a year | Method: SPIN CALL BUDGET Length: 45 minutes, set by deal band. Expect a finance sign-off and a light security review. Both add calendar time that has nothing to do with your product. Airtime target: you about 18 min, them about 27 min (40% you). SPIN puts most of the airtime on implication questions, which are long to answer, so your share falls naturally. Talk share is an output, not a lever. The number you can actually fix is your longest single monologue, so cap it near 90 seconds. SEQUENCE 1. FRAME (4 min) [GATHER] Before we start: what made you take this call? I want to spend the time on that rather than on my slides. [GATHER] How long have you been running director of claims operations here, and what did you inherit when you took it on? 2. SITUATION, KEPT DELIBERATELY SHORT (5 min) [GATHER] I read that you handle this through a mix of internal process and tooling. What did I get wrong? [GATHER] Who touches the workflow between the start and the point where it is considered done? [GATHER] How often does that run, and what is the volume in a normal month? 3. PROBLEM (10 min) [GATHER] Where does that process break down most often? [REALISE] What part of it do people work around rather than use as designed? [REALISE] If I asked the person doing this every day what the worst part of it is, what would they say? 4. IMPLICATION, WHERE THE CALL IS WON (14 min) [REALISE] When that happens, what does it stop from happening elsewhere? [REALISE] How much time does the team lose to it in a month, roughly, and whose time is it? [REALISE] What does the customer or the internal stakeholder see while that is being sorted out? 5. NEED-PAYOFF, IN THEIR WORDS (7 min) [REALISE] If that were not happening at all, what would you do with the time it frees up? [REALISE] What would it be worth to have that fixed before your next peak period? [REALISE] Who else would notice if this changed, and what would they say about it? 6. CLOSE (5 min) [GATHER] What has to happen between now and a decision, in order? [GATHER] What would make this not worth continuing, from your side? QUESTION MIX 16 questions: 8 gathering, 8 creating realisation (50% realisation). Gathering questions fill your CRM. Realisation questions make the buyer assemble a cost they had been absorbing in pieces. Below about 40 percent realisation you are running an interview; the buyer leaves knowing nothing new and nothing changes. Any gathering question answerable from their website should be converted into a confirmation before the call: say what you found, ask what it is missing. QUALIFICATION BAR (SPIN is not a qualification framework, so add one) 1. The buyer stated the problem themselves, unprompted, in their own vocabulary. 2. The buyer, not you, put a number or a duration on the consequence. 3. The buyer said out loud what solving it would be worth. If you said it for them, the need-payoff stage did not happen. 4. A date exists with an owner and a consequence behind it. 5. You know what happened the last time they bought something this size. Three or fewer of these and the call produced rapport rather than qualification. COMPELLING EVENT QUESTIONS 1. Is there a date by which this has to be sorted, and what happens on the day after it if it is not? 2. Who is accountable if that date passes? Name the person, not the team. 3. Is anything expiring, renewing, being switched off or being audited in the next two quarters? 4. What is on your objectives this year that this would contribute to, and when is that reviewed? A real event has three parts: a date, an owner, and a consequence for a named person. Two out of three is a preference with a deadline attached. BUYING PROCESS QUESTIONS (7 at this deal size) 1. What happened the last time you bought something of this size? Walk me through it from first conversation to signature. 2. Above what figure does this need an approval you cannot give? 3. Who else has to be comfortable with this, even if they do not sign anything? 4. How long is the gap between a verbal yes and a signed contract here, in practice rather than in policy? 5. Is there a security or IT review, and can it run in parallel rather than after commercials? 6. Does procurement get involved, and at what point do they normally enter? 7. Who redlines the contract, and how long did that take last time? Ask historically, not hypothetically. The described process is the official one; the last purchase tells you the real one.

Frequently asked questions

How many questions should a discovery call have?

Scale it with the length of the call rather than picking a number. One widely quoted vendor analysis of recorded sales calls put the productive range at roughly eleven to fourteen questions, which fits a thirty to forty-five minute conversation and would be far too many for a fifteen-minute qualification call. This tool sets the count from the deal band, because a five-thousand-dollar deal that takes twenty minutes and a two-hundred-thousand-dollar deal that takes an hour cannot share a question list without one of them being wrong.

What is the difference between a gathering question and a realisation question?

A gathering question moves information to you and its answer goes in the CRM. A realisation question moves information to the buyer and its answer usually starts with a pause, because it forces them to assemble something that was previously stored in several people's heads separately. You need both, but a call made only of gathering questions is an interview, and buyers do not change anything because a vendor now understands them. The output tags each question so you can see the mix rather than assume it.

Does the methodology really change the questions, or just the labels?

It changes the stage sequence, the weighting of minutes, and which section carries most of the call. SPIN pushes situation questions into pre-call research and spends most of the time on implication. MEDDIC weights towards decision process and metrics and produces a scored gap list at the end. Challenger opens with a reframe before any question is asked, so the questions that follow are testing whether the claim landed rather than collecting facts. The tool changes the structure accordingly, not just the headings.

Why does the tool ask what happened last time they bought something?

Because people describe the official buying process and then follow the real one. The official version leaves out the finance meeting that only sits monthly, the security review that adds three weeks, and the approval threshold above which an unfamiliar director gets involved. Asking historically rather than hypothetically surfaces all three, and it is a comfortable question to answer because it is about the past rather than about your deal.

What counts as a real compelling event?

A date, an owner, and a consequence that lands on a specific person if the date passes. All three, or it is a preference with a deadline written next to it. The fastest test is to ask what happens the day after. A contract terminating, an audit, a system being switched off and a regulatory deadline all pass. Wanting better reporting this year does not, however sincerely it is meant.

Related free tools

Lead Scoring Model Builder Sales
B2B Customer Avatar Builder Sales
Sales Objection Handler Sales
Cold Calling Scriptwriter for recruitment agencies Outbound
AI Lead Magnet Ideator Content

Some links on this site are affiliate links, which means Hacking Demand may earn a commission if you buy through them at no extra cost to you. This does not influence which tools are listed. The tools on this page are free and have no affiliate relationship of any kind.