Sales Discovery Questions: 25 Questions That Reveal Real Problems

Most lists of sales discovery questions create the same failure they promise to solve. A rep opens a document with 40 prompts, asks them in order, and leaves with a completed checklist but no shared understanding of the buyer's problem. The prospect feels interviewed. The CRM looks full. The deal is still vague.

A discovery call has a more demanding job: turn an initial reason to talk into a diagnosis both sides can test. You need to understand what changed, how work happens now, where the process breaks, what the failure costs, and what evidence would justify action. That means asking fewer questions, following the answer, and recording facts separately from assumptions.

The 25 questions below are organized around that job. They are a bank, not a script. Choose the questions that fit the situation, ask one at a time, and let the buyer's answer determine your next move.

What a discovery question should produce

In an August 2026 episode of Sales Talk for CEOs, sales practitioner Becc Holland described discovery as a diagnostic process rather than an opportunity to present the product. Her comparison was a medical screening: the professional uses a finite set of questions to reveal a condition the person may not yet recognize. The point is not to manufacture pain. It is to help someone examine observable behavior and its consequences. Listen to the episode.

A useful discovery question should produce at least one of four things:

• A fact: what happens, who does it, how often, and with which constraint.
• A contrast: the gap between the current state and the expected state.
• A consequence: what the gap changes for the team or the business.
• A test: the evidence needed to decide whether change is worthwhile.

Questions that only invite agreement do none of these. "Would more qualified pipeline help?" is easy to answer and useless for diagnosis. "What happens to an account after a rep decides it is qualified?" exposes a real process you can examine together.

The TRACE discovery framework

Use TRACE to move from curiosity to evidence without jumping straight to a product pitch. Each stage answers a different question.

1. Trigger: why is this conversation happening now?
2. Routine: how does the work happen today?
3. Anomaly: where does reality differ from the expected process?
4. Consequence: what does that difference change for the business?
5. Evidence: what would justify a decision and a next step?

The sequence matters, but it is not rigid. If the buyer gives you a consequence while describing the routine, explore it. Then summarize what you heard and verify it before moving on. Discovery is a conversation with a map, not a form with required fields.

25 sales discovery questions, organized by purpose

1. Trigger: understand why the conversation exists

Start with the event or idea that earned attention. This keeps the call grounded in the buyer's context instead of your agenda.

1. What changed that made this worth looking at now?
2. What in our message or conversation made you curious?
3. Why has this become a priority now rather than six months ago?
4. Who first raised the issue internally, and what did they notice?
5. If nothing had changed, would we still be having this conversation?

Question two is especially useful in outbound. Holland argues that a curious prospect should not immediately receive a BANT checklist. First ask what created the curiosity. Their answer tells you which hypothesis deserves testing.

2. Routine: reconstruct how work happens today

Buyers often describe a problem in labels: "lead quality," "conversion," or "productivity." Labels hide the mechanism. Ask for the sequence of actions instead.

6. Walk me through what happens from the first signal to the first sales action.
7. Who touches the process, and what decision does each person make?
8. What information does a rep have before deciding whom to contact?
9. Where is that information recorded, checked, or handed off?
10. What does a normal week look like when the process works as intended?

"Walk me through" is powerful because it invites a description rather than a verdict. Listen for handoffs, manual work, missing evidence, and decisions made by habit. Do not label any of them as a problem until the buyer does or the evidence supports it.

3. Anomaly: find where the process departs from expectation

Now compare the routine with the outcome the team expects. The goal is not to collect complaints. It is to locate a repeatable failure or constraint.

11. Where does this process most often slow down or lose accuracy?
12. Which decision is hardest to make with the information available?
13. What do your strongest reps do differently from the rest of the team?
14. Which accounts get worked even though they rarely become opportunities?
15. What must someone check manually before they trust the result?

Notice the difference between "Do you waste time on bad leads?" and question fourteen. The first assumes the answer and invites politeness. The second asks the buyer to identify a pattern. If they cannot, you do not yet have evidence for the problem.

4. Consequence: connect the anomaly to business impact

A problem is not automatically a priority. Establish what changes when it occurs and who feels the effect. Use the buyer's numbers when they have them. Do not supply a benchmark just to make the case sound urgent.

16. When that happens, what work has to be repeated or abandoned?
17. Who feels the effect first: the rep, the manager, marketing, or the buyer?
18. Which metric moves when this process underperforms?
19. What does the team postpone because this work consumes its attention?
20. If the process stayed the same for another two quarters, what would that change?

Ask for a number only when the buyer can connect it to a real measurement. "We think this costs time" is a hypothesis. "Four reps each spend two hours every Monday cleaning the list" is a fact you can validate. Keep those categories separate.

5. Evidence: define a credible next step

Finish by agreeing on what must be true before either side invests more time. This is more useful than forcing budget and timing questions into the first ten minutes.

21. What would you need to see to believe this problem is worth changing?
22. Which part of the process should we test first, and on what sample?
23. Who else can confirm the workflow and its impact?
24. What could make this change fail even if the product works?
25. What is the smallest next step that would reduce the uncertainty?

A sound next step is an evidence-gathering action: review a sample, verify a workflow, involve the process owner, or test a defined result. "Book a demo" is an activity. It becomes a useful next step only when everyone knows what the demo must prove.

How to ask without turning discovery into an interrogation

The quality of a question depends on what you do after the buyer answers. Use a four-part loop:

1. Ask one finite question. Do not stack three prompts into one sentence.
2. Follow the evidence. Ask for an example, a sequence, or the last occurrence.
3. Summarize without upgrading the claim. Keep uncertainty visible.
4. Verify. Ask whether your summary is accurate or missing context.

For example: "You said reps check three sources before contacting an account. The part I am not clear on is whether they do that for every account or only strategic ones. Which is it?" The follow-up is specific, neutral, and easy to correct.

Avoid performing expertise by explaining every answer back to the buyer. Holland's broader point in the podcast is that sellers need domain knowledge so they can ask useful questions and recognize an important signal. That knowledge should improve your listening. It should not occupy the call.

Five weak questions and better replacements

Weak question	Why it fails	Better question
What keeps you up at night?	Generic, theatrical, and detached from the reason for the call.	What changed that made this worth examining now?
Do you struggle with lead quality?	Leading question that contains the seller's conclusion.	How does a rep decide an account is ready for outreach?
What is your budget?	Asks for a commitment before establishing the value of change.	How does your team fund changes to this process once the case is clear?
Are you the decision-maker?	Reduces a complex decision to hierarchy and can make the contact defensive.	Who needs to validate the process, the risk, and the commercial decision?
Would automation save time?	Invites an obvious yes without measuring the current work.	Which step repeats most often, and how is it handled today?

What to record after the call

A discovery note should preserve the quality of the evidence, not just the conclusion. Record six fields while the conversation is fresh:

• Fact: an observable statement the buyer confirmed.
• Hypothesis: an interpretation that still needs testing.
• Exact language: the buyer's short phrase for the problem or desired state.
• Impact: the metric, work, delay, or risk connected to the issue.
• Owner: the person who operates or validates the process.
• Next evidence: what the next meeting, sample, or test must establish.

This structure prevents a common CRM failure: a rep writes "pain confirmed" when the buyer only agreed that the topic was interesting. It also gives managers something concrete to coach. They can examine the evidence behind the opportunity instead of debating how confident the rep sounds.

Where qualification ends and discovery begins

Lead qualification and discovery solve related but different problems. Qualification asks whether the company and contact deserve attention. It can use firmographics, role, current signals, and an ideal customer profile before the first call. Discovery asks whether a meaningful problem exists, how it works, and whether change is worth pursuing.

Do not make discovery repair poor targeting. If the account has no plausible fit, a brilliant question will not create one. Do not make pre-call scoring pretend it knows the buyer's internal process either. Observable account signals can justify attention; the conversation supplies the diagnosis.

Use AI to prepare and preserve evidence, not invent it

AI can make discovery preparation faster. It can summarize public account signals, propose questions tied to a known workflow, compare call notes with required evidence, and identify unanswered questions. It is also capable of turning a reasonable inference into a confident false fact.

Set three operating rules:

• Every account-specific question must be tied to a source or marked as a hypothesis.
• The call summary must preserve uncertainty and the speaker behind each claim.
• A human verifies the evidence before changing qualification, forecast, or next-step status.

Lead Scorer helps teams research accounts, score fit, and prepare outreach with the supporting sources visible for review. That makes the first conversation more relevant. The rep still owns the diagnosis.

A 10-minute preparation checklist

Before your next discovery call, prepare one page:

1. Write the verified trigger for the conversation.
2. List two facts about the account and label every other assumption.
3. Choose one TRACE question for each stage.
4. Write the process you expect, then plan to let the buyer correct it.
5. Define the evidence that would justify another meeting.

During the call, abandon the list when an answer deserves a follow-up. At the end, summarize the facts, the unresolved hypothesis, and the smallest useful next step. The goal is not to ask 25 questions. It is to leave both sides with a clearer and more accurate view of the problem than they had at the start.