--- name: discovery-call-prep description: Prepares for a discovery call with research and a tailored question set, then after the call summarizes it, scores qualification, and proposes the deal update. Use when the user asks to "prep for my discovery call", "give me discovery questions for [company]", or afterward "summarize my discovery call", "update the deal from this call". license: MIT compatibility: Requires a connected CRM and web search; a transcript for the post-call mode. Drafts and proposes updates — no auto-writes. metadata: author: Builder.org version: 2.0.0 persona: Strategic Steve rave: R4 A3 V3 E3 category: crm --- # Discovery Call Prep & Post-Call Summary ## What it does Covers both sides of a discovery call. Before: research on the prospect and a question set built around a stated hypothesis, so the call tests something instead of wandering. After: a summary, a qualification read, and a proposed deal update for confirmation. One skill, the whole call. ## When to use - Before: "prep for my discovery call with [company]", "discovery questions for this prospect". - After: "summarize my discovery call", "update the deal from this call". - Not for general meeting follow-ups on non-discovery calls; that is `meeting-follow-up-coach`. Not for research with no call booked; that is `prospect-intelligence-deep-dive`. ## Inputs Prep mode: the company or contact; the rest comes from your CRM (any your assistant can read) and the web. Post-call mode: the transcript or notes, plus the deal if not obvious. Optional in both: your qualification framework (defaults to MEDDIC-lite: pain, metric, decision process, decision maker, timeline). ## The bar - Prep is built on a hypothesis, stated in one sentence. Questions without a hypothesis are a survey, and prospects can feel the difference. - Ten questions maximum, ranked, with the three must-asks marked. A 25-question list is a script for a call nobody enjoys. - Questions are phrased to earn stories, not yes/no answers: "walk me through what happens when..." beats "do you have a problem with...". - Post-call: every qualification claim cites the transcript. What the prospect did not say is listed as open, not assumed. - Proposed deal updates are proposals. Stage moves and field writes wait for confirmation, always. ## Workflow 1. **Research the prospect.** Contact role and background, company firmographics, recent signals from the CRM and the web. Note how they arrived (inbound form, referral, cold outbound), because it changes what the call must establish. 2. **Form the hypothesis.** One sentence: what pain probably brought them here and what would make this deal real. The call exists to confirm or kill it. 3. **Build the question set.** Up to ten, grouped: situation (their world today), pain (where it hurts and what it costs), process (how they buy, who decides), and timing (why now, what deadline). Rank them; mark the three must-asks. Tailor at least half to the research; generic questions waste the research you just did. 4. **Set the target next step.** The specific advance to ask for if the call goes well (a scoping session, an intro to the decision maker), so the close of the call is decided before it starts. 5. **After the call, summarize.** From the transcript: needs, pains with their costs, budget and timeline signals, objections, and who else was named. Verbatim quotes for the two or three lines that matter most. 6. **Score qualification and propose the update.** Each framework element: confirmed (with the quote), partial, or open. Then the proposed CRM update: stage, next step with a date, fields worth writing, and the follow-up email draft. Present for confirmation; write nothing until confirmed. ## Output ``` ## Discovery prep · [Company] **Context:** [who, how they arrived, key signals] **Hypothesis:** [one sentence to test] **Must-asks:** 1. [question] 2. [question] 3. [question] **Also, if time:** [remaining questions, grouped] **Target next step:** [the advance to ask for] ## Post-call · [Company] **Summary:** [4-6 bullets, quotes where it matters] **Qualification ([framework]):** [element: confirmed/partial/open, evidence] **Proposed update (confirm):** stage [X to Y] · next step [action, date] · [fields] **Follow-up email draft:** [subject + body] ``` ## Worked example ``` Fictional example. Never reuse these names or numbers in real output. ## Discovery prep · Juniper Point Financial **Context:** Nadia Brooks, Director of Client Services, booked from the pricing page after downloading the compliance guide. 40-person wealth management firm, two offices, hiring a client onboarding coordinator (careers page, checked today). **Hypothesis:** client onboarding is manual enough that compliance review is the bottleneck, and the new-hire req means they are about to solve it with headcount instead of software. **Must-asks:** 1. "Walk me through what happens between a client signing and their first portfolio review today." 2. "You're hiring an onboarding coordinator; what will that person spend their first month fixing?" 3. "When onboarding slips, who notices first, you or the client?" **Also, if time:** Situation: how many onboardings per month; which systems touch the process. Process: who besides Nadia evaluates tools; what compliance sign-off requires. Timing: what makes this quarter the moment; what happens if nothing changes. **Target next step:** a 45-minute process-mapping session with Nadia and their compliance lead within two weeks. ``` ## Never - Never bring more than ten questions, or bury the must-asks in the middle of the list. - Never ask what research already answered; convert those into confirmations ("sounds like the Riverside win doubled your volume, is that right?"). - Never report a qualification element as confirmed without transcript evidence. - Never auto-apply a stage change, field write, or task. Propose, confirm, then write exactly that. ## Recovery - Thin research (private company, quiet contact): say the hypothesis is low-confidence and add one extra situation question to rebuild it live in the first five minutes. - Post-call with sketchy notes instead of a transcript: summarize only what the notes support, mark the qualification mostly open, and list the specific gaps to close by email. - The call revealed the hypothesis was wrong: say so plainly in the summary; a killed hypothesis with a new one attached is a successful discovery call. - No framework given and the deal is small: score against pain, decision maker, and timing only, and say the lighter bar was used and why.