How to run a discovery call
Run a discovery call that leads to a sale: book every yes within 2 days, open with the time and the goal, ask 12 questions about their problem and let them talk more than half the call, check 4 signs they fit, say your price, book a dated next step before you hang up, send the recap within 24 hours, and count your calls every Sunday.
ByDinesh BypillaReviewed byMurali SidLast checked 21 min read
After weeks of outreach, someone finally said yes to a call. This guide shows how to run a discovery call: book it within two days, ask 12 questions about their problem, say your price, and leave with a dated next step. Most founders spend the call showing their product, hear "sounds great, send me something", send it, and never hear back.
Do these three things
About 45 minutes a call, with the preparation and the recap. Keep your outreach going so the calls keep coming, and judge your calls after 20, not after 3.
Show the three actions in fullHide the full list
- Today: make a free booking page with 20-minute slots and two questions, and write a one-page call sheet with 12 questions about the buyer's problem and 4 checks that tell you if they fit.
- Every day: hold every discovery call you can book, within 2 days of the yes. Let them talk more than half the call, say your price, book a dated next step before you hang up, and send the recap within 24 hours.
- Every Sunday: count calls booked, held, fit, next steps and wins in one sheet. After 20 calls, fix the first step where most calls stop.
Get help doing these
ChatGPT or Claude reads this guide and works through it with you.
1. Book every discovery call within two days and send the agenda
Reply to every yes the same day with two times in the next two days and a booking link. Then send a short agenda, and a reminder on the morning of the call. A call booked a week away is a call that often does not happen.
The yes comes from your outreach: a reply to one of your 100 daily personal messages, a LinkedIn conversation, a phone call or an enquiry from your site. Answer it the same day, in the same thread. Do not send only a link: give two real times first, so they can answer with one word.
Set up the booking page once:
- 1
Make the booking page
In Google Calendar, click Create, then Appointment schedule. Set 20-minute slots, add a 10-minute buffer after each one, and pick the hours you keep for calls. Click Next, then Save. It is free with any Google account. Many founders keep calls to the afternoon, so the mornings stay for outreach. Cal.com's free plan works the same way if you prefer it.
- 2
Add two questions to the booking form
"What made you want to talk now?" and "What is your website?" The first gives you the reason for the call. The second is what you read when you prepare.
- 3
Keep the call to the time you asked for
Most outreach messages ask for 15 or 20 minutes. 20 is enough for 12 questions if you do not pitch. If they want to keep talking, ask before you run over.
When they book, send the confirmation below. On the morning of the call, send one line: "Still good for 2pm? Here's the link: [link]." Do not count on the booking tool to send it. Write it yourself. It takes 30 seconds.
Subject: Our call on [day]: 3 thingsHi [first name],Thanks for booking. Here's what I'd like to cover in our [20] minutes: How you handle [the problem] today, and what it costs you. Whether what I do fits. I'll tell you straight if it doesn't. If it fits, what a next step would look like. If someone else would have a say in this, they are welcome to join.[Your name] [Call link]
- Tool: Google Calendar's appointment schedule (free), Gmail and a Google Meet link.
- Time: 30 minutes to set up, then 3 minutes for each booking.
- You will have: a booking page, a confirmation you can send in a minute, and every call booked within 2 days of the yes.
- It worked if: at least 8 of every 10 booked calls happen. If fewer do, book closer to the yes and always send the morning reminder.
- Common mistake: booking the call for next week because your calendar looks full. Interest fades in days. The other mistake is a long agenda. Three lines are enough.
See it done: Quillfox's confirmation
Quillfox is a made-up one-person agency. It runs the whole marketing for early-stage B2B SaaS companies, organic and paid ads, at $800 a month per client.
To: Priya, founder of a 6-person SaaS company
Subject: Our call on Wednesday: 3 things
Hi Priya,
Thanks for booking. You wrote that most visitors leave your homepage without starting a trial. Here's what I'd like to cover in our 20 minutes:
- How people find you today, and what the missing trials cost you.
- Whether what I do fits. I'll tell you straight if it doesn't.
- If it fits, what a next step would look like.
If your co-founder would have a say in this, he's welcome to join.
Jess [Call link]
Why it works: It repeats the founder's own reason for the call, promises an honest answer on fit, and invites the person who decides.
2. Prepare for 10 minutes before each discovery call
Spend 10 minutes before every call on their website, their LinkedIn and their answer on the booking form. Write three guesses about their problem on your call sheet. You go in knowing what to ask, and you never ask what their company does.
Buyers now come to the first call with most of their research done. In 6sense's 2025 survey of nearly 4,000 business buyers, they first spoke to a seller about 61% of the way through their buying journey, and the seller they spoke to first won about 8 in 10 deals. "So what does your company do?" tells them you did not look.
Paste their homepage text and their booking answer into ChatGPT or Claude. Ask for a 5-line summary: what they sell, to whom, how big they look, and what might be going wrong for them with the problem you fix. Then write your three guesses in your own words on the call sheet. Check anything the AI says against their site. It guesses too.
Name | Company | Role | Date How they found us: Their reason for the call (booking form): Guess 1 about their problem: Guess 2: Guess 3: The one thing I most need to learn on this call:
- Tool: their website and LinkedIn, ChatGPT or Claude, and your call sheet in Google Docs or Sheets.
- Time: 10 minutes a call.
- You will have: a call sheet with the top filled in before every call.
- It worked if: you can say your three guesses out loud in 20 seconds, and at least one of them comes up on the call.
- Common mistake: spending an hour on research. Ten minutes is enough. The call is where you learn the rest.
3. Start the discovery call with the time, the goal and a question about recording
Open every call the same way: thank them, check the time, say what you will cover, ask if you may record, and ask how they found you. It takes 2 minutes, and both of you know where the call is going.
Ask before you record, every time. In 12 US states, including California, Florida and Illinois, everyone on a call must agree to a recording, and when the two of you are in different places, the stricter rule usually applies. If they say no, take notes by hand.
A recording is worth asking for. Fathom's free plan records and transcribes calls on Zoom, Google Meet and Teams with no limit, and gives full AI summaries for the first 5 calls a month. You need the transcript for step 6 and step 7. A free Google Meet call between two people can run for up to 24 hours; with 3 or more people, it stops at 60 minutes.
Thanks for making the time, [first name]. We have [20] minutes. Does that still work?Here's what I'd suggest. I'd like to ask about [the problem] and how you handle it today. Then, if it sounds like I can help, I'll tell you how I'd do it and what it costs. At the end we decide together if a next step makes sense, and it's completely fine if it doesn't.Is it OK if I record this, so I can listen instead of taking notes?Before we start: how did you find us?
"How did you find us?" is self-reported attribution. Write the answer on your sheet, then in your CRM if you use one. It is how you learn which channel brings deals, and whether people now find you through ChatGPT.
- Tool: Google Meet or Zoom, and Fathom's free plan.
- Time: 2 minutes.
- You will have: an agreed plan for the call, a yes or no on recording, and where they found you.
- It worked if: they say yes to the plan in the first 2 minutes, and their source is on your sheet.
- Common mistake: opening with 10 minutes about your company, your story and your clients. They booked the call to talk about their problem. Your story can wait until they ask. If you are a technical founder who dreads this part, read the opening word for word until it feels normal.
4. Ask 12 questions about their problem and let them talk more than half the call
Ask 12 questions about their problem: what happens today, the last time it went wrong, what it cost, why now, and who decides. Spread them through the call, and ask "What happened then?" after each answer. Do not show or explain your product until they have described the problem.
The questions follow the order buyers think in. A sales training firm, Winning by Design, calls it SPICED: the situation, the pain, its impact, the critical event (a date that forces a decision) and how the decision gets made. You do not need the name. You need the questions.
Today What made you want to talk now? How do you handle [the problem] today? Walk me through it. What have you looked at or tried already? What did you like and not like about it? The last time 4. Tell me about the last time [the problem] happened. What happened? 5. How often does that happen?What it costs 6. What did it cost you: time, money or customers? 7. What happens if nothing changes in the next 3 months?Why now 8. Is there a date or event this needs to be ready for?Who decides 9. Who else would have a say in this? What will they want to know? 10. What have you paid for something like this before? 11. How do you usually buy something like this: card, invoice or contract? 12. If this works, what would you need to see to get started?
Questions 4 to 6 matter most. "The last time" gets you a real story, not a polite guess about the future. That is the rule from customer interviews, and it holds on a sales call too. When they give you a number, write it down in their words: "We lose about 40 trials a month."
How much to talk. Gong studied 326,000 sales calls in 2025. On average, sellers talked 60% of the time. On deals they won, they talked 57%; on deals they lost, 62%. Those were all kinds of sales calls, not only discovery calls. On a discovery call you mostly ask, so aim to talk less than half. Gong also found that sellers who won asked 15 to 16 questions, and those who lost asked about 20. An older Gong study of discovery calls, first published in 2017, found the best results at 11 to 14 targeted questions, spread through the call. 12 good questions with follow-ups beat 20 in a row.
When to talk about your product. Only after question 8, and only about the part that fixes the problem they named. If they ask "How does yours work?" earlier, answer in two sentences and come back: "I'll show you properly in a minute. Can I ask one more thing first?" If they want to see the product, book a demo as the next step instead of turning this call into one.
Note the tools they named in question 3. It is the cheapest competitor research you will ever do, and their words for the problem belong in your cold emails and your homepage headline.
If you run a B2B business: ask question 9 early if the person on the call is not the owner or the head of the team. If more than two people have to agree, see how to sell to bigger companies. If they mention a formal tender, ask for the date it opens: that is a different job, responding to an RFP.
If you run an agency: ask "Who did this for you before, and what happened?" after question 3. Their answer tells you what they will check you for. Ask also what exactly they want done each month. A scope you never asked about is how an agency ends up doing $4,000 of work for $2,000, so write their list down for your scope of work.
If you run a D2C brand: your discovery calls are with shop buyers, distributors and companies that want to order in bulk. Swap questions 10 to 12 for: "How many units would a first order be?", "What margin do you need on a product like this?" and "When do you next choose new products?" More on getting into shops.
- Tool: your call sheet with the 12 questions, and the Fathom transcript.
- Time: about 15 minutes of the call.
- You will have: their problem, what it costs and their date, written in their own words.
- It worked if: you talked less than half the call, asked 11 to 16 questions, and your sheet has a number for the cost and a date. Check the talk share after the call: paste the transcript into ChatGPT or Claude and ask what share of the words each person spoke.
- Common mistake: pitching the moment they mention a problem. The call turns into a presentation, and you leave without knowing what it costs them. The other mistake is reading all 12 questions like a form, one after another.
See it done: Jess's call sheet after the call with Priya
Priya, founder, 6-person SaaS company
| Question | Priya's answer |
|---|---|
| What made you want to talk now? | "We get about 1,500 visits a month and only 12 trials. Our investor asked about it." |
| What have you tried? | "A freelancer wrote 4 blog posts last year. Nothing changed." |
| The last time it went wrong | "We posted on LinkedIn for a month. Lots of likes from other founders, 2 trials." |
| What it costs | "A trial turns into a paying customer about 1 time in 4. So every 4 lost trials is $1,200 a year." |
| Date | "I want 30 trials a month by January, before we raise." |
| Who else decides | "My co-founder signs off on anything over $1,500 a month." |
| Paid before | "The freelancer was $900 a month." |
Why it works: Every answer is in Priya's own words and has a number, so the price and the next step have something to stand on.
5. Check if they fit and say your price before the call ends
Before the last 5 minutes, check four things on your sheet: a problem that costs them, a date, the person who decides, and a price they can live with. Then say your price out loud. A buyer who hears the price first in a proposal a week later is the one who goes quiet.
The four checks:
- The problem costs them. It happened in the last 3 months, and they can say what it cost.
- There is a date. A launch, a season, a funding round, a contract ending. No date usually means no decision.
- The person who decides is on the call, or will join the next step.
- The price is OK. They did not say no to your range.
The rule: 3 or 4 ticks, they fit, so book the next step (step 6). 2 or fewer, say so: "From what you've told me, I don't think I'm the right fit yet, because [reason]. Can I check back in 3 months?" Then ask if they know one person who has the problem now. Most first clients come from introductions like this. An honest no saves you a proposal nobody reads, and it often gets you an introduction.
Say the price before the call ends, whatever the ticks say. Use the sentence from your pricing work:
For [a company like yours / an order like yours], most [clients / orders] are between $[low] and $[high]. For you I'd start with [option], which is $[price] [a month / in total]. Does that fit what you had in mind?
Then stop talking and wait for the answer. If they say "That's more than we thought", ask "More than what?" Their answer tells you whether to offer the smaller option or to show the value again, using the cost they gave you in step 4.
If you run a SaaS: say your plan and its price, and how long the free trial is. Your price is usually on your pricing page, so say it anyway: "The Team plan is $[price] a month. Does that fit?" See how to set your SaaS price.
If you run an agency: name the package you would start with, as in your price list: "For a company like yours, I'd start with Core, which is $[X] a month."
- Tool: your call sheet.
- Time: 3 minutes.
- You will have: a fit mark (3 or 4 ticks, or not), and their answer to your price, on every call.
- It worked if: every call ends with the price said out loud and the ticks counted. After 20 calls, at least half the calls you held should fit. If fewer do, the problem is who you message, not the call.
- Common mistake: hiding the price until the proposal "so they see the value first". You spend a week on a proposal for someone who could never pay it.
6. Book the next step before you hang up, then send a recap within 24 hours
If they fit, agree one next step with a date while you are both still on the call, and send the calendar invite before you hang up. Within 24 hours, send a short recap in their own words, with the price and the next step. Then follow up on day 3, day 7 and day 14 if they go quiet.
The next step depends on what you sell:
| If they need… | The next step is… |
|---|---|
| to see your product working | a product demo of the part that fixes their problem |
| a written price for their boss or co-founder | a call to walk through your proposal, with the person who decides |
| to try it themselves | setting up their account together, on this call or the next (trial users who start with you are more likely to stay) |
| to hold the product | a sample sent today and a call the day it arrives |
| something to send around | a one-page summary or a short sales deck, and a call with the others |
Say it like this: "The next step would be [step]. Shall we put it in now? How's [day] at [time]?" Then send the invite while you are still talking.
Close on the call or book the next step? If all four ticks are there and they ask "How do we start?", take the yes. Tell them exactly what you will send today: the agreement, the invoice or the account invite. Then onboard them the same week, so the first result comes fast. Otherwise, book the next step. Do not push for a yes they have not offered.
After the call, paste the transcript or your notes into ChatGPT or Claude with the template below, and ask for a draft. Check every number and every promise yourself. Send it within 24 hours. If they fit, send the written price or quote the same day too.
Subject: [their company]: notes from our callHi [first name],Thanks for the call. Here's what I heard, so you can tell me if I got it wrong: [Their problem, in their words.] [What it costs them, with their number.] [Their date.] What I'd do: [one or two lines]. As I said on the call, that's $[price] [a month / in total].Next step: [step] on [day, date, time], with [who]. The invite is in your calendar.[Your name]
If they do not answer, follow up in the same thread on day 3, day 7 and day 14, each time with one new thing: a short answer to a question they asked, a case study of a similar client, or a date that is getting close for them. After day 14, move them to a monthly nurture email.
Log every call in your sheet the same day: date, name, source, ticks, price said, next step and its date. Once the sheet gets slow, move to a CRM. Every call is a deal in your sales pipeline, even the ones that say no.
- Tool: Google Calendar, Gmail, ChatGPT or Claude to draft, and your Google Sheet.
- Time: 3 minutes on the call, then about 15 minutes for the recap and the log.
- You will have: for every call that fits, an invite in both calendars and a recap sent within 24 hours.
- It worked if: at least half the calls that fit end with a dated next step, and all of your recaps go out within 24 hours.
- Common mistake: ending with "I'll send you something" and no date. That is how most "sounds great" calls go quiet. The other mistake is a recap that talks about you. Use their words.
See it done: Quillfox's recap
To: Priya, founder of a 6-person SaaS company
Subject: Loomtally: notes from our call
Hi Priya,
Thanks for the call. Here's what I heard, so you can tell me if I got it wrong:
- About 1,500 visits a month bring only 12 trials.
- About 1 trial in 4 becomes a paying customer, so every 4 lost trials cost you about $1,200 a year.
- You want 30 trials a month by January, before you raise.
What I'd do: posts from your own LinkedIn profile, plus comparison pages for the two bigger tools buyers check before they try you. As I said on the call, I'd start with Core, which is $2,100 a month.
Next step: we walk through the three options on Monday 12 October at 3pm, with your co-founder. The invite is in your calendar. I'll send the options today.
Jess
Why it works: It proves Jess listened, repeats the price she said on the call, and names a date with the co-founder who signs off.
7. Count your discovery calls every Sunday and fix the first step where they stop
Every Sunday, fill in one row: calls booked, held, fit, price said, next steps and wins. After 20 calls, find the first step where most calls stop, and fix only that step.
Week | Calls booked | Held | Fit (3–4 ticks) | Price said | Next step booked | Won | Where the calls came from
Then open 2 of this week's transcripts and check two numbers: the share of words you spoke and how many questions you asked. Paste each transcript into ChatGPT or Claude and ask for both.
After 20 calls, fix the first step that fails its line:
- Held: fewer than 8 of 10 booked calls happened. Book within 2 days and send the morning reminder (step 1).
- Fit: fewer than half the calls you held fit. Change who you message, not the call.
- Next step: fewer than half the calls that fit got a dated next step. Listen to 2 recordings. Look for talking more than half the time, a missing price, or no date asked for (steps 4 to 6).
- Won: 10 next steps and no win. The call is working; fix the proposal or the demo.
Keep the calls coming while you fix. The calls come from your outreach: 100 personal messages a day to a list that keeps growing, 3 follow-ups for each person, and introductions from partners and clients. At about 3 conversations for every 100 messages, that is a few calls a day, so 20 calls take weeks, not months. Ask every client you win for an introduction and a testimonial too.
- Tool: Google Sheets, and the Fathom transcripts.
- Time: 30 minutes every Sunday.
- You will have: one row a week, and one thing to fix next week.
- It worked if: you can name the step where most of your calls stop, and next week's number for that step goes up.
- Common mistake: judging after 3 calls, or by how the call felt. "Great call" without a next step is a lost call.
See it done: Quillfox after 20 calls
Three weeks of calls
| Booked | Held | Fit | Price said | Next step booked | Won |
|---|---|---|---|---|---|
| 20 | 17 | 7 | 17 | 5 | 2 |
The decision: 17 of 20 calls happened, so booking works. Only 7 of 17 fit, under half. 6 of the 10 that did not fit had no paying customers yet and no budget. So Jess adds "has paying customers" to her lead list rule and keeps the call as it is.
Why it works: Jess finds the leak before the call, so she changes who she messages instead of rewriting a call that works.
When your calls fit and most get a next step, the work moves on to the demo and the proposal. Once a month, look at the whole picture in a marketing review.
Your discovery call checklist
0 of 13 done
Work through this list with an AI assistant
It drafts the work and keeps track of what is done.
Frequently asked questions
- How long should a discovery call be?
20 minutes is enough for most first calls with a small business, if you ask 12 questions and do not pitch. Keep to the time your message asked for, often 15 or 20 minutes. Book 30 only when more than one person joins or the deal is large. If they want to keep going, ask first: "We're at 20 minutes. Do you have 10 more?"
- What is the difference between a discovery call and a demo?
On a discovery call you ask about their problem, check if they fit and say your price. On a demo you show how your product fixes the problem they told you about. Keep them apart: a demo before discovery shows every feature to someone whose problem you do not know yet.
- Should I talk about price on a discovery call?
Yes. Say a range or your starting price before the call ends, and ask "Does that fit?" If it does not, you have saved a week on a proposal. If it does, the proposal holds no surprise. Saying your own price gets a real reaction. "What's your budget?" often gets a guess.
- Which framework should I use: BANT, MEDDIC or SPICED?
You do not need one by name. The 12 questions in this guide cover what they all ask: the situation, the problem, what it costs, the date, who decides and the money. SPICED is close to this order. Look at the others only if you start selling to large companies, where more people decide.
- What do I do if they go quiet after a good call?
Follow up in the same thread on day 3, day 7 and day 14, each time with one new thing, such as an answer to their question or a similar client's result. Next time, book the next step on the call and say your price before you hang up. Most calls go quiet when they end with "I'll send you something".
Sources
- 1.Mastering the talk-to-listen ratio in sales calls — Gong, 21 August 2025, updated 4 March 2026
- 2.Effective strategies for successful sales discovery calls — Gong, 2017, updated 4 March 2026
- 3.The 2025 B2B Buyer Experience Report — 6sense, 2025, checked 5 October 2026
- 4.The SPICED framework — Winning by Design, checked 5 October 2026
- 5.Create an appointment schedule — Google Calendar Help, checked 5 October 2026
- 6.Pricing — Cal.com, checked 5 October 2026
- 7.Pricing — Fathom, checked 5 October 2026
- 8.Learn what requirements you need to use Google Meet — Google Meet Help, checked 5 October 2026
- 9.Call Recording Consent Laws by State (2026 Guide) — Viirtue, 29 July 2026
Next article to read
How to do a product demoOnce you understand their problem, show them how the product solves it.Read it nextRelated guides
- How to do customer interviewsDo customer interviews when you have almost no customers: 7 questions about the last time the problem happened, 100 messages a day to book 20-minute calls, a notes sheet in their exact words, and a Sunday count that puts those words into your marketing.
- How to price a B2B product or serviceB2BPrice your B2B product or service when you do all the selling: work out your floor price, collect 20 real prices, check the value to the buyer, quote the same day in three options, and change the price only after 20 quotes.
- How to write a proposal that winsWrite a proposal that wins: say your price range and agree a decision date on the call, send a one-page proposal with three options the same day to everyone who decides, follow up on day 3, 7 and 14, and change one part after every 10 proposals.
- How to move from spreadsheets to a CRMMove from spreadsheets to a CRM in one afternoon, for free: put everyone who replied into a five-stage pipeline, log every reply the same day, and review your deals every Sunday.
Guides that link here
- How to market your B2B businessB2B
- How to price a B2B product or serviceB2B
- How to do customer interviews
- How to get your first B2B clientsB2B
- How to get clients for a consulting business without referralsB2B
- How to follow up on a cold email
- How to do LinkedIn outreach
- How to cold call past the gatekeeperB2B
- How to do a product demo
- How to make a sales deckB2B
- How to sell to enterprise clients
- How to respond to an RFPB2B
- How to host a webinar
- How to get your company recommended by ChatGPT as a supplierB2B
- How to track AI traffic and ChatGPT mentions
- How to create a B2B lead list for free
- How to find a decision maker's email addressB2B
- How to get replies to cold emails
- How to warm up an email domain for cold email
- How to write a cold email
- How to get booked as a podcast guest
- How to repurpose content across platforms
- How to start a YouTube channel for business
- How to use AI for content creation
- How to write a case study
- How to get customers online for free
- How to get people to try your product
- How to get your first client as a marketing agencyAgency
- How to build a waitlist before you launch
- How to launch a B2B productB2B
- How much to spend on marketing as a B2B companyB2B
- How to choose marketing channels for an agencyAgency
- How to create an ideal customer profile
- How to do a competitor analysis
- How to fit marketing into your week as a solo founder
- How to market your marketing agencyAgency
- How to market your product as a technical founderSaaS
- How to price a SaaS productSaaS
- How to price digital marketing servicesAgency
- How to set a marketing budget for your agencyAgency
- How to validate a business idea without spending money
- How to write a value proposition
- How to do account-based marketingB2B
- How to find distributors for your product
- How to get clients from networking eventsB2B
- How to get meetings from a trade showB2B
- How to get reviewed on Clutch
- How to get your product into retail storesD2C
- How to run a co-marketing campaign
- How to set up referral partnerships
- How to design a pricing page
- How to move from spreadsheets to a CRM
- How to write a proposal that wins
- How to measure SEO success
- How to set up a Google Business Profile
- How to write competitor comparison and alternative pages
- How to build a community around your brand
- How to build in public on XSaaS
- How to find online communities where your buyers talk
- How to grow on LinkedIn as a founder
- How to write a LinkedIn profile that brings in clients
- Instagram post ideas for a marketing agencyAgency
- LinkedIn post ideas for B2B companiesB2B
Cite this page
Dinesh Bypilla (2026). How to run a discovery call. Scout7 Academy. https://scout7.ai/academy/sales/how-to-run-a-discovery-call
Written by Dinesh Bypilla, Agentic systems engineer at Scout7. Reviewed by Murali Sid. Last checked 5 October 2026.
Quote any part of this guide with a link back to this page. © 2026 Scout7.
Markdown version of this page: https://scout7.ai/academy/sales/how-to-run-a-discovery-call.md