Discovery calls often go wrong in a very boring way.
The buyer says, "We need better visibility." The rep hears a pain point, writes visibility in the notes, and starts thinking about the dashboard demo.
Reasonable. Also a little too fast.
"Better visibility" is not much of a problem yet. It is the safe, compressed answer people give when they have explained the same issue in three planning meetings and two vendor calls. Nobody has said what is hard to see. Nobody has said who makes a worse decision because of it. And nobody has asked what happened the last time the team got it wrong.
The first question finds the subject. The second question starts the discovery.
The first answer is usually the safe one
Ask about a business problem and you will often get the version that already fits on a slide: improve efficiency, increase conversion, reduce costs, get more visibility.
Nothing dishonest there. The answer is just compressed to the point of being almost useless.
The tempting move is to pounce on the first useful noun and pitch against it. Resist that for another minute.
"Visibility into what?" keeps the conversation open. "Can you walk me through the last time that caused a problem?" is better. Now there is a day, a person, and probably an awkward meeting somewhere in the answer.
Earn the right to ask
A discovery call should not begin with questions Google could answer. Headcount, office locations, recent funding, the buyer's job title. Do the homework first.
Amanda Georgeoff put it cleanly in a LinkedIn Sales article: "The right to ask a question is earned in sales."
That does not mean arriving with a 40-slide account plan. One useful observation is enough. The point is to open with a view, not ask the buyer to give you a company tour.
"I saw you added 18 field reps this year. Teams at that stage often find that meeting context arrives in the CRM days late, if it arrives at all. How are you handling that now?"
You may be wrong. Fine. A useful hypothesis gives the buyer something specific to correct. A blank question makes them do the work of deciding what your meeting is about.
Salesforce reports that 87% of business buyers expect sales reps to act as trusted advisors. "So, tell me about your business" is not trusted-advisor territory.
Use the second-question rule
The rule is simple: when a buyer gives you a tidy answer, do not reward it with a pitch. Ask once more.
Look for an example, a consequence, or a choice. Any one of the three will tell you more than another feature question.
| First answer | Useful second question | What it may reveal |
|---|---|---|
| "Forecasting is difficult." | "What happened in the last forecast call that made that obvious?" | Missing evidence, late updates, or management pressure. |
| "Reps hate CRM admin." | "Which update is most likely to be skipped after a customer meeting?" | The exact workflow that is broken. |
| "We need this soon." | "What changes if it is still unsolved next quarter?" | Whether urgency is real or merely polite. |
| "The team needs more visibility." | "Who is making a worse decision today because they cannot see it?" | The person, decision, and consequence involved. |
The follow-up has to grow out of what the buyer just said. If your next question would be identical no matter how they answered, you are not really listening. You are reading a checklist with eye contact.
Five follow-ups worth keeping
1. Ask for the last real example
"When did this last happen?" pulls a fuzzy problem into the real world. Suddenly there is a trigger, an owner, a workaround, and a result.
If the buyer cannot remember an example, do not rescue the answer for them. The problem may not be urgent. That is useful information.
2. Find the point where the process breaks
"The process is manual" is a description, not a diagnosis. Where does the work wait? What gets copied twice? Which step disappears when Friday gets busy?
Try: "Walk me through what happens after the rep leaves the meeting." Then listen for the handoff nobody owns, the field nobody fills, or the spreadsheet held together by one heroic employee and a terrifying number of tabs.
3. Follow the consequence downstream
Every operational problem sends a bill somewhere. Sometimes it is lost time. Sometimes it is a forecast nobody trusts, a delayed launch, another meeting, or a marketing campaign built on half a CRM.
Ask, "Who feels that first?" or "What decision gets harder because of it?" An annoyed user is worth understanding. A compromised business decision is usually what gets funded.
4. Separate importance from urgency
A buyer can care deeply about a problem and still do nothing this year. LinkedIn's roundup of discovery questions from standout sellers draws a useful line: how important is the problem, and how important is it to solve now?
Very different questions. The gap between them has swallowed plenty of optimistic forecasts.
5. Ask how the decision actually gets made
Please do not ask, "Are you the decision-maker?" It makes a capable champion feel small and gives everybody else a reason to say yes.
Ask about the last purchase instead: "Can you walk me through the last time your team bought something like this?" The answer usually contains finance, legal, an executive who appears late, and the company's private definition of "soon."
Retire the questions that sound profound but go nowhere
Some questions look excellent on a sales-training slide and feel ridiculous when said to another adult.
"If you could wave a magic wand..." is the famous one. It invites a broad answer with no constraint, owner, or decision attached. LinkedIn's discovery guide puts it on the avoid list for exactly that reason.
The same problem appears in questions like:
- "What keeps you up at night?" Too intimate before you have earned it, and usually rewarded with a rehearsed answer anyway.
- "What is your biggest challenge?" Too wide. Give the buyer an area to react to.
- "Would saving time be valuable?" A yes-shaped question that teaches you nothing.
- "What would 10x growth mean for the business?" A dramatic hypothetical when you still do not understand what happens on Tuesday afternoon.
Plain language wins. Ask about work that happened, decisions that are coming, and consequences the buyer recognises without your help.
How many discovery questions should you ask?
There is a wonderfully tempting numeric answer. Gong says its team analysed 519,291 sales conversations and found that successful discovery calls often included 11 to 14 targeted questions, spread through the conversation rather than fired in one burst.
Useful benchmark. Terrible quota.
A buyer who gives one detailed answer and gets three thoughtful follow-ups may tell you more than a rep who completes 14 boxes. If you need 30 disconnected questions to understand the deal, the buyer is not on a discovery call. They are helping you fill out a form.
A good discovery call has structure. It should not feel structured.
The number is not the interesting part of Gong's analysis. The interesting part is the shape: a small set of real problems, several layers of follow-up, and enough space for the buyer to think.
Do not flatten the answer when the call ends
One final problem: great discovery can still die in the CRM five minutes later.
The rep hears a five-minute story about an unreliable handoff, a furious regional manager, and a forecast that moved by $800,000.
The CRM gets: "Needs better reporting."
Neat. Almost useless.
A colleague who missed the call should still be able to understand why the deal exists. That means keeping:
- Current process: what happens now, in the buyer's words.
- Break point: the step where work slows, disappears, or gets repeated.
- Evidence: the most recent example, not just the category of pain.
- Consequence: the decision, metric, person, or customer affected.
- Urgency: why the team may act now, and what happens if it does not.
- Decision path: who joins, what they care about, and how a purchase is approved.
- Next step: owner, date, and what each side agreed to bring.
That is the difference between a CRM full of activity and one that explains the pipeline. LogicNotes for teams turns the actual conversation into structured notes, tasks, and CRM-ready context before a useful five-minute answer gets reduced to three vague words.
Quick answers
What is the single best discovery question?
There is no universal one. "Can you walk me through the last time that happened?" is a strong default because it moves the conversation from opinion to evidence.
Should discovery questions be open-ended?
Mostly. Open questions invite context. A precise closed question can still confirm a date, owner, or fact. The problem is not yes-or-no questions; it is leading the buyer towards the answer you hoped to hear.
When should a rep disqualify the opportunity?
When there is no meaningful problem, consequence, plausible decision path, or reason to act. A pleasant conversation is not automatically a deal.
Research note: This guide draws on discovery-call guidance from LinkedIn Sales Solutions, Salesforce, HubSpot, and Gong. LogicNotes added the examples, practical follow-up patterns, and CRM workflow.
Keep the answer behind the answer.
LogicNotes turns sales conversations into summaries, action items, and CRM-ready context, so the next call starts with what the buyer actually meant.
See LogicNotes for teams