Discovery Call Checklist
Seven sections, from the ten minutes before the call to the hour after it. Tick through it live in this page, copy it into your notes, or use the prompt at the bottom to rewrite the whole thing for one specific call.
No email, no signup. Word opens in Google Docs, the spreadsheet opens in Sheets, and the whole template is on this page as text either way.
The template
Before the call: 10 minutes, not 40
Enough to sound informed and to have one specific opening question. Past ten minutes you are procrastinating.
The first five minutes: set the frame
You are buying permission to ask hard questions later. Skip this and every question after it feels like an interrogation.
“Before I show you anything, I would rather understand what is actually going on, so I do not waste 20 minutes on the wrong half of the product. Is it all right if I ask questions for the first stretch?”
Current state: how it works today
You cannot quantify a pain until you know what the process is. Reps skip this and end up selling against an imaginary status quo.
Pain and impact: four questions, in order
Each one builds on the last answer. Asked out of order they all land flat, and question four asked first gets you a polite non-answer.
Decision, money, and process
Qualification is not a separate call. All five of these belong in discovery, not in a follow-up you have to earn.
The last five minutes: land a next step
A call that ends with “I will send some information over” is a call that ended without a next step.
“So what I heard is [pain], which is costing you [number], and the reason it matters this quarter is [driver]. What did I get wrong?”
Within an hour of hanging up
The detail you did not write down is gone by the afternoon.
Filled in: a real-shaped call, start to finish
Every company, person, and number below is invented. It is here to show the level of specificity the template is asking for.
Discovery call: Wexford Logistics, VP Data Platform, 34 minutes
The checkpoints on the left, what was actually said on the right. Notice how much of the right-hand column is a direct quote.
Scroll the table sideways for the remaining columns.
| Checkpoint | What was actually said |
|---|---|
| Hypothesis going in | They posted two data reliability engineer roles in six weeks. Guess: something broke publicly and nobody owns detection. |
| What prompted the call | “Our CFO opened a revenue dashboard in the board meeting and the number was nine hours stale. I have been asked to make sure that cannot happen again.” |
| Current state | 1,400 tables in Snowflake. dbt tests on about 120 of them. Alerts go to a Slack channel with 40 people in it and no owner. Last process change was in January, when they added the channel. |
| Pain, question 1 | “We find out about broken data from whoever opens the dashboard first.” |
| Pain, question 2 | “The board asked for a weekly revenue view. That is the first time anyone outside the company sees these numbers, so being wrong is now visible.” |
| Pain, question 3, the metric | Mean time to detection. Currently around nine hours, measured informally. Nobody reports it. The CFO now asks about it monthly. |
| Pain, question 4, why now | “The board pack goes out on the 5th. If it is wrong twice, this becomes a headcount conversation instead of a tooling one.” |
| Cost of doing nothing | Estimated two analyst days per incident, roughly two incidents a month, plus the board risk they would not put a number on. |
| Who else agrees | Her, the Head of Analytics (owns the BI contract, may resist), and the CFO signs anything over $50k. |
| Decision criteria, verbatim | “It has to cover raw tables, not just the modelled ones, and it cannot need my team to write more tests.” |
| Last purchase like this | Bought their BI suite in 2024. Took eleven weeks, security review was the long pole, procurement needs a SOC 2 report up front. |
| Alternatives | The BI vendor's bundled module, and “hiring two more people and writing more dbt tests”. |
| Next step agreed | Technical session Thursday 3pm with her and one platform engineer. She is inviting the Head of Analytics herself. |
| What we got that we did not have | The nine-hour detection number, and the name of the Head of Analytics. |
| Question I should have asked | Whether the CFO has actually seen a detection-time number, or is just asking for one. |
Rewrite it for your next call
A generic checklist gets you through the call. A checklist written for the company you are about to meet gets you the second one. Give it the prospect and your own site, and it rebuilds all seven sections around that call.
Your details
Opens claude.ai or chatgpt.com in a new tab with your prompt ready.
You are an elite B2B sales strategist. Turn the generic discovery call checklist below into a call-specific one for the call described here. Research the prospect's company before you write. **Prospect company:** ___ **Who I am meeting:** ___ **My company website:** ___ **What I already know:** ___ **Research first.** Read their website, recent announcements, and any open job postings. Job postings are the most honest signal of what a team is struggling with. Note anything that suggests a trigger event in the last two quarters. **Produce a checklist with these seven sections, every item rewritten for this specific call:** 1. **Before the call.** Five research items, each naming the specific thing to look at for this company rather than a generic instruction. End with the single hypothesis this call is testing, stated as a falsifiable sentence. 2. **The first five minutes.** How to frame the call for this persona, and the exact opening question. Include one verbatim line for getting permission to ask questions before presenting. 3. **Current state.** Four questions that map how this specific process works at a company of this shape. Name the tools they most likely already use. 4. **Pain and impact.** Four questions in a fixed order: surface the challenge, find the business driver behind it, quantify the metric that is suffering, then establish why now. Customise each one to this role and industry. For each, list two or three likely answers and the follow-up probe for each. 5. **Decision, money, and process.** Five qualification questions phrased as this persona would actually answer them. Avoid "who is the decision maker". 6. **The last five minutes.** How to play back what was heard, and two or three next steps that would be realistic for this deal shape, most ambitious first. 7. **Within an hour of hanging up.** What to log, and what to send. **Also produce:** - A "signals to listen for" list: three answers that mean this is a real deal, and three that mean it is a tire-kicker. - The two questions most likely to be skipped under time pressure, and which sections to cut instead. **Rules:** - Conversational phrasing. Every question must sound natural said out loud, not read off a form. - No question that can be answered yes or no, except the permission question in section 2. - Keep the whole thing to one page. A checklist that does not fit on a screen does not get used on a call.
How to use this discovery call checklist
A checklist is not a script. Read it before the call, keep it open during the call, and expect to use maybe two thirds of it. The sections are ordered the way a good call runs, but the order inside each section is yours: a prospect who opens by describing their pain has just answered section four, and continuing to section three would be filling in a form at them.
The four pain questions have to run in order
This is the part reps get wrong most often. Question four, “why now”, is the one everyone wants to ask, and asked first it gets a shrug. It only works after the prospect has said their challenge out loud, connected it to something happening in the business, and named the metric that is suffering. By then they have talked themselves into urgency, and question four just asks them to say the date.
Write down quotes, not summaries
“Concerned about data quality” is your language. “Our CFO opened a dashboard in the board meeting and the number was nine hours stale” is theirs, and it is the sentence that will still be moving the deal three weeks later when you quote it back to a stakeholder who was not on the call. The worked example above is mostly quotation marks for exactly this reason.
Qualification belongs in discovery
There is a habit of treating discovery as the nice call and qualification as the awkward one that happens later. That gets you a pipeline full of deals with no economic buyer. Five questions in section five cover most of what you need, and the highest-yield one is asking what the last purchase like this looked like internally. One answer gives you the paper process, the timeline, and the blocker.
Judge the call on the next step, not the rapport
Calls that feel great and end with “send me some information” are the most common failure in discovery. Before you hang up, you want a date, a named attendee list, and one thing you did not have when you dialled in. If you cannot get those, the honest read is that you did not find pain worth their calendar time, and the fix is upstream in section four.
The checklist is open in another tab
Every rep who has run a discovery call knows the feeling of remembering question three on the drive home. A checklist you have to look away from the prospect to read is competing with the conversation. Backdrop pushes the next question onto the screen as the call moves, based on what was just said, so the sequence holds even when the call does not go the way it was planned.
See how Backdrop worksFrequently Asked Questions
What is a discovery call checklist?
- A discovery call checklist is a structured list of what to research, ask, confirm, and log across a first sales call. It exists so the parts reps drop under time pressure, quantifying the pain and landing a dated next step, survive a call that runs off the plan.
What should you cover on a discovery call?
- How the process works today, what is going wrong with it, which business driver makes that a priority now, the metric that is suffering, who else has to agree, what has to be true for them to say yes, what else they are considering, and a specific next step with a date on it.
How long should a discovery call be?
- Thirty minutes is normal and enough. Book the last five for playback and the next step, and treat the middle twenty as the working part of the call. If you need sixty minutes to get through this checklist, you are presenting during discovery rather than asking.
How many questions should you ask on a discovery call?
- Fewer than the checklist contains. Aim for eight to twelve real questions with follow-up probes, not thirty asked at speed. A prospect should be talking for most of the call, and every question you skip in favour of listening to a longer answer is usually the right trade.
What is the most important part of a discovery call?
- Turning an expressed complaint into a number. A prospect saying data quality is a problem gives you nothing to build a business case on. The same prospect saying detection takes nine hours and costs two analyst days per incident gives you a deal, a stakeholder, and a reason to move now.