Guides
Use 35 sales discovery questions to uncover context, impact, stakeholders, requirements, priority, and a responsible next step, then practice them with Lira.
Published
Sales discovery questions should help a rep understand what changed, how the buyer operates today, what the problem affects, who owns the decision, what evidence matters, and whether there is a responsible next step. The goal is not to ask all 35 questions. It is to choose the 11 to 14 that fit the buyer and let the answers determine the follow-up.
See how Lira helps reps practice discovery conversations
A useful question changes what the seller understands or does next. Salesforce defines a discovery call as an early sales conversation for understanding fit, needs, and motivations. Gong's analysis of more than 519,000 B2B sales calls found that asking 11 to 14 targeted questions correlated with the strongest results in its dataset, while front-loading questions made the call feel like an interrogation. The practical lesson is to prepare more questions than you will use and earn each follow-up from the buyer's answer.
The strongest discovery question is open enough to reveal context and specific enough to move the decision forward. It should not smuggle in a product claim, ask for information the rep could have researched, or force the buyer to accept the seller's framing.
| Question group | What it should reveal | Weak substitute |
|---|---|---|
| Context and change | Trigger, current process, prior attempts | Tell me your pain points |
| Business impact | Frequency, consequence, metric, desired outcome | How big is the opportunity? |
| People and decision | Owners, users, approvers, skeptics | Are you the decision-maker? |
| Requirements | Must-haves, constraints, workflow, proof | Which features do you want? |
| Priority and next step | Risk, resources, timeline, mutual action | Can I send a proposal? |

Treat this as a question bank, not a script. A rep should enter the call with a hypothesis and a route through the five groups, then skip questions the buyer has already answered and follow evidence that changes the evaluation.
Start with the buyer's reality before diagnosing it. The rep's first job is to understand the triggerTriggerA trigger is an event or condition that starts, resumes, changes, or stops an automated workflow. and the current operating model without pretending the account is a blank page.
This is a better opener than asking for a generic list of priorities. It gives the buyer room to name the trigger, urgency, and internal context in their own words.
The current process is the baseline for every later claim. A rep who skips it cannot distinguish a real gap from a workflowWorkflowA workflow is a defined sequence of steps, decisions, actions, delays, and outcomes used to complete a business process. the buyer already likes.
Discovery should not assume everything is broken. Preserving what works makes the conversation more credible and reveals the buyer's nonnegotiables.
Ask for a moment, handoff, or decision where the process fails. Concrete breakdowns are more useful than a broad statement that the system is inefficient.
This question separates the person attending the call from the people carrying the operational cost. It also points toward users, champions, and affected teams.
Past attempts reveal constraints, internal beliefs, and change fatigue. A strong rep treats a failed attempt as evidence, not as an invitation to criticize another vendor.
The barrier may be ownership, process, data, budget, adoption, or a missing capability. The answer should shape qualification and the next step, not merely create another talking point.
Impact questions should connect an operating problem to a measurable consequence or a clearly owned outcome. They should not manufacture a financial case the buyer cannot defend.
Follow the consequence through the workflow. The first symptom is rarely the full impact, so the rep should ask what work, customer experience, risk, or revenue outcome changes next.
Frequency helps distinguish an occasional nuisance from a repeatable operating problem. Do not force a number the buyer cannot support; a reliable range is more useful than false precision.
Existing metrics show how the buyer already thinks about value. If the team does not measure it, the rep should say so plainly rather than inventing an ROI model.
A project without an owned success measure is vulnerable to enthusiasm without accountability. The best answer connects the initiative to a metric the buyer's team actually reviews.
Manual cleanup, rework, escalations, duplicate entry, and missed handoffs often carry more weight than the visible problem. This question tests whether the issue spreads beyond one team.
A future-state answer clarifies expectations without asking the buyer to accept a seller's promise. The rep should listen for both the result and the operating behavior behind it.
A date is meaningful only when it connects to a launch, target, contract, hiring plan, or other real event. Manufactured urgency is not qualification.

Complex decisions are systems of judgment, not org-chart trivia. Map who experiences the problem, who changes the workflow, who evaluates risk, and who can commit resources.
Ownership reveals who can validate the current state and who will be accountable after a purchase. If nobody owns it, the deal may need internal alignment before a product evaluation.
The buyer and the user are often different people. A recommendation that ignores the user's workflow can win a meeting and still lose adoption.
Ask about approval as a process, not a hunt for a single executive. Finance, security, legal, procurement, operations, and business leaders may each own a different part of the decision.
A skeptical stakeholder is useful information, not an obstacle to route around. Their concerns should shape proof, participation, and the next conversation.
Change cost belongs in discovery. The more teams that must alter tools or habits, the more the evaluation should test implementation and ownership.
Different stakeholders rarely use one scorecard. This question helps the rep separate financial, technical, operational, and user criteria instead of presenting one generic pitch.
Past decisions expose the real path more reliably than a hypothetical process. The rep should use the answer to reduce surprises, not to pressure the buyer into a timeline.
Requirements matter only when they connect to the buyer's workflow, constraints, and evidence standard. A feature checklist without that context rewards surface similarity.
A disciplined evaluation separates requirements from preferences. Without that distinction, every feature looks equally important and the buyer cannot explain the tradeoff later.
Budget, policy, security, data, geography, accessibility, staffing, and timing can all be legitimate boundaries. Discover them before proposing a path that cannot survive review.
Integration questions should begin with the workflow and data movement, not a logo checklist. A named system matters only if the buyer can explain what must pass between it and the proposed solution.
Some processes are stable for good reasons. This question surfaces trusted controls, customer expectations, and internal practices that a seller should preserve rather than bulldoze.
A product can fit the use case and still fail the operating model. Resourcing belongs in qualification because implementation work does not disappear after signature.
The status quo, an internal build, a process change, and another vendor can all be real alternatives. A useful discovery call makes those options explicit without pretending the seller is neutral about the result.
Ask whether the buyer needs a pilot, reference, security review, workflow test, business case, or stakeholder workshop. The answer defines what proof must do and prevents a demo from becoming theater.
Priority is visible in ownership, resources, consequences, and a real next action. The rep should make it easy for the buyer to say no or not yet when those signals are absent.
Implementation disruption, adoption, vendor dependence, data exposure, and a weak result are different risks. The rep should name and test them rather than answering a concern the buyer did not raise.
Inaction has a cost only when the buyer can describe it. This question tests whether the project is truly a priority without turning the call into an urgency script.
Stalled work often reveals missing ownership, unclear success criteria, competing priorities, or an approval gap. The answer should change the deal plan.
Budget should be discussed directly and respectfully. A range, owner, and approval path are more useful than forcing a precise number before the buyer has enough information.
A contract date, launch, hiring plan, planning cycle, or compliance milestone is evidence. A seller's quarter end is not the buyer's deadline.
The right next meeting should close an evidence gap. Add people because their judgment is needed, not because a larger calendar invite makes the deal look healthier.
End discovery with named owners, evidence, and a date. A vague promise to follow up is not a next step; it is an opportunity to lose the context the call just created.
Do not select questions by copying a methodology into the calendar invite. Select them from what is already known, what remains uncertain, and which uncertainty could change fit, value, risk, or the next step.
Review the account, role, recent trigger, existing relationship, and prior notes. Remove any question the rep can answer credibly without the buyer. Nothing drains trust faster than asking a busy person to repeat public facts.
Prepare a view of the likely problem, impact, and stakeholder map, then turn each into a testable question. A hypothesis makes the rep sharper. Treating it as a conclusion makes the rep deaf.
Pick eight to ten core questions across the five groups and reserve several follow-ups for the answers that matter. The research cited above supports a focused 11-to-14-question conversation, not a 35-item interrogation.
| Buyer answer | Rep response | Reason |
|---|---|---|
| Specific and evidenced | Summarize, confirm, then move forward | Do not punish clarity with repetitive probing |
| Broad or abstract | Ask for an example, range, owner, or recent instance | Make the answer operational |
| Uncertain | Separate what is known from what needs validation | Uncertainty is information |
| Contradicts the hypothesis | Acknowledge it and change the route | Discovery should be allowed to disconfirm fit |
| Reveals risk or a new stakeholder | Pause the pitch and map the implication | The deal plan has changed |
Explore Lira for repeatable AI sales roleplay
Lira can turn a discovery standard into a spoken practice conversation. Current Lira product guidance says teams can build scenarios around their products, buyers, and conversations, let reps speak with an AI counterpart in the browser, configure skills and objectives, review feedback, and repeat the attempt. That makes Lira a practice and readiness layer. Managers still own the qualification standard, company facts, scoring judgment, and live deal decisions.
Define the buyer role, trigger, current approach, information the buyer knows, details they will reveal only after a relevant question, legitimate constraints, and the condition for a next step. Random resistance is not realistic discovery practice.
| Skill | Observable evidence | Critical miss |
|---|---|---|
| Preparation | Uses known context and tests a relevant hypothesis | Asks for public facts |
| Question choice | Moves from context to impact, decision, and next step | Follows a fixed list regardless of answers |
| Listening | Summarizes accurately and uses the answer in the next question | Repeats or ignores the buyer's point |
| Depth | Asks for examples, owners, ranges, or consequences | Accepts vague pain language as proof |
| Judgment | Disconfirms weak fit and avoids unsupported claims | Forces every answer toward the product |
| Next step | Confirms owners, evidence, and timing | Ends with an unowned follow-up |
A second attempt should have a purpose: replace a leading opener, follow an answer more naturally, quantify without cornering the buyer, map a skeptical stakeholder, or confirm a mutual next step. Repeating the whole call without a focus creates activity, not deliberate practice.
Use the sales role play scripts guide for scenario openings and curveballs, the sales training program guide for rollout and manager calibration, and the AI sales coaching guide for the boundary between practice feedback and accountable manager judgment. Those pages cover the training system; this article owns the discovery-question bank and its practice standard.

A 35-question call is usually a questionnaire wearing a headset. Use the bank to prepare, then let relevance and answer quality determine the route.
Questions such as 'Would it help if you could automate that?' hide a pitch inside the discovery process. Ask what the buyer needs, then decide honestly whether Lira or any other product belongs in the answer.
A real problem may still lack ownership, resources, timing, or executive attention. Qualification improves when the rep can distinguish an acknowledged issue from a funded decision.
Do not convert a guess into ROI. Record the buyer's measure, range, assumptions, and source, then make uncertainty visible in any business case.
A calendar invitation is not progress by itself. The next step should name the question being resolved, who must participate, what each side will bring, and when a decision can be made.
Sales discovery questions are questions used to understand a prospect's current situation, business impact, stakeholders, requirements, decision process, priority, and appropriate next step. They help a seller test fit and plan evidence before prescribing a solution.
There is no universal count, but Gong reported that 11 to 14 targeted questions correlated with the best outcomes in its analysis of more than 519,000 B2B sales calls. Use that as a directional benchmark, not a quota. The right count depends on answer depth and call scope.
Discovery builds an evidence-based understanding of the buyer and the problem. Qualification uses that evidence to decide whether the opportunity, solution, resources, and next step make sense. They overlap, but discovery should be allowed to prove that the deal is not a fit.
Use a structured route, not a verbatim script. Prepare the opening, core question groups, transitions, proof boundaries, and next-step language, then adapt to the buyer's answers. Scripts are useful scaffolding and poor substitutes for listening.
No. AI roleplay can make spoken practice, configured feedback, and repeat attempts more available. Managers still own product truth, qualification criteriaQualification criteriaQualification criteria are the required signals an agent must collect before a lead can be booked, transferred, quoted, or routed., scorecard calibration, disputed feedback, deal context, and readiness decisions.
Lira lets teams build company-specific conversation scenarios, have reps practice aloud with an AI counterpart in the browser, configure skills and objectives, review feedback, and repeat. It is not a CRMCRMA CRM is the system used to manage leads, contacts, accounts, opportunities, activity, ownership, and follow-up., live-call recorder, autonomous cold dialer, outbound calling product, or all-in-one enablement suite.
Practice your discovery standard with Lira
Product capabilities and boundaries were checked against current first-party Lira pages immediately before publication. External research is attributed to the organization that published it. The question bank, selection rules, and scorecard are editorial recommendations, not universal causal claims or a claim of hands-on product testing.
The best discovery call is not the one with the longest question list. It is the one that turns preparation into a relevant dialogue, makes uncertainty visible, and leaves both sides clearer about whether and how to proceed. Practice that judgment, not just the words.