The single highest-leverage hour in any sales cycle is the first discovery call. Everything downstream — the demo, the proposal, the negotiation, the close — is downstream of how well that first call went.
So it's worth pausing on this: most reps run discovery badly, and the reason isn't talent. It's that discovery is treated as a skill (something you have or don't) rather than a structure (something you execute). The founders and AEs who close consistently aren't more charismatic. They're running the same 7-part structure on every call, and they got there by writing it down.
Here's the structure. Steal it.
Why discovery fails
Walk into any early-stage sales org and watch three discovery calls. You'll see three completely different conversations. Different opening, different question order, different depth on pain, different closing. The win rates correlate exactly with the rep who happens to have a natural feel for the rhythm.
This is the artisanal sales motion, and it has two problems:
- It doesn't transfer. New AEs can't learn from it because there's no pattern to copy.
- It doesn't improve. Without a structure, you can't tell whether a call went badly because the prospect wasn't a fit or because the call wasn't run well.
The fix is not "make the reps better at discovery." The fix is to make discovery a structured deliverable — the same way you wouldn't run a board meeting without an agenda.
The 7-part discovery structure
1. Frame (2 minutes)
Open with the agenda, the time box, and the outcome. Specifically:
"I've got us for 30 minutes. My goal is to figure out whether what we do is a fit for your situation — and if it's not, I'll tell you directly. Here's what I'd love to cover: 5 minutes on your role and what you're working on, 15 minutes on the specific challenge that put us on this call, and the last 10 we'll spend on whether and how we might work together. Sound good?"
Three things this does: sets a time box (respect signals competence), establishes that you'll disqualify (signals confidence), and explicitly frames a decision conversation at the end (eliminates "I'll need to think about it").
2. Situation — 5 minutes, not 15 (5 minutes)
The most common mistake: spending 20 minutes on company background, team structure, current tools. Most of that is on their LinkedIn or their website.
You need exactly three things in this section:
- What's your role and what are you accountable for? (the personal stake)
- What does your team look like and what's your current stack for [problem space]? (the situation)
- What changed recently? (the trigger — almost always the most important question on the call)
If you spend more than 5 minutes here, you've broken the structure. Move on.
3. Problem — and make it specific (8 minutes)
This is the section that separates good discovery from great discovery. The question is never "what are you trying to solve?" — that produces generic answers. The question is:
"Walk me through the last time this came up. What happened?"
You want a specific story. A real Tuesday afternoon. A specific person. A specific Slack thread. Generic problems don't close — they become "we'll revisit next quarter." Specific problems close.
While they tell the story, you're listening for three things: who else was involved (multi-threading signal), what they tried first (existing tooling), and what the cost of the failure was (Implication territory — see section 4).
4. Implication — make the cost real (5 minutes)
Most reps stop at problem. They feel good about it ("they have a clear problem!") and pivot to demo. Do not pivot to demo. A problem without quantified Implication is not a budget-funded purchase, it's a "nice to have."
Your job is to surface the cost. Examples:
- "When that happens, what's the downstream impact?"
- "How often does it happen? How many people does it affect?"
- "What's it costing you — in time, in revenue, in someone's quarterly target?"
If the prospect can't articulate a cost, the deal will die in procurement. Better to find that out now than three weeks into a proposal cycle.
5. Decision process — before, not after (5 minutes)
This is the section reps systematically skip. Don't. Ask:
"If you and I decided today that this was a fit, what would the next steps look like on your side? Who else gets pulled in? How does something like this get approved here?"
You're looking for: how many people, how much budget authority your contact has, what the typical timeline for a decision like this is at their org, and whether legal/security/procurement are involved.
The reason this lives in discovery, not in the proposal stage: by the time you're sending a proposal, the cost of finding out it needs board approval is two weeks of stalled deal. Find out now.
6. Disqualify — explicitly (2 minutes)
The most powerful 2 minutes of a discovery call:
"Based on what you've told me, here's where I think we might be a fit, and here's where we might not be. We're really strong for [specific case]. We tend to be a poor fit for [specific case]. From what you've described, you sound more like the first — does that match your read?"
This does three things at once: it tells the prospect you're not pitching them, it surfaces objections they were hiding, and it lets them self-confirm their own buying motivation in their own words. (Read that last one twice. That's the magic.)
7. Mutual close — and a real next step (3 minutes)
Never end a discovery call with "I'll send you some materials and we'll regroup." Never. The next step is:
"Here's what I'd suggest: I take what we just talked about, I put together a 20-minute demo focused on [the specific story they told you in section 3], and I bring in our [solutions engineer / customer who looks like you / etc.]. Can we put it on the calendar before we hang up?"
Calendar on the call. Not "I'll send some times." Not "let me check with my team." On. The. Call. If the prospect won't put it on the calendar, the deal isn't real, and you just learned that in 30 minutes instead of 30 days.
How to make this stick across a team
A 7-part structure on a Notion page is worthless. Here's how it actually becomes the default for your team:
- Record every discovery call. Non-negotiable.
- Score each call against the 7 sections. Did the rep frame? Did they keep Situation to 5 minutes? Did they get a specific story in Problem? Did they quantify Implication? Did they map Decision process? Did they disqualify? Did they calendar?
- Pick the weakest section per rep, and drill it for 2 weeks. Not "be better at discovery." Specifically: "for the next 10 calls, you will end every Problem section with a quantified Implication question."
- Track the section-level scores over time. This is where compounding improvement happens.
If you've read our post on why win rates don't move, this is the same principle: decompose the call, coach one dimension, measure the delta.
This is exactly the work AscentGTM does in the background — every discovery call scored against Qualification + Diagnostic Discovery dimensions automatically, the section-by-section gap surfaced per rep, the coaching drill suggested. The 7-part structure becomes a shared operating system instead of a Notion doc nobody opens.
TL;DR
- Discovery is the highest-leverage hour in your sales process. Reps run it badly because it's treated as art, not structure.
- The 7 parts: Frame, Situation (5 min), Problem (specific story), Implication (quantify cost), Decision process, Disqualify explicitly, Mutual close on the calendar
- Specific Tuesday-afternoon stories close. Generic problems don't.
- Skipping "Decision process" is the #1 reason proposals die three weeks later
- Make the structure transferable: record, score against the 7 sections, drill the weakest one per rep
Discovery is where the deal is won or lost. Run it like a system, not like a vibe.