Running the Deal: discovery that surfaces the truth, and proposals that close

Your team already has the discovery-system, post-call-system, and proposal-and-pricing playbooks — the worked recipes for planning a discovery call, capturing it into structured signal, and turning that into a tailored proposal, step by step. This module is the layer above the recipe. It’s where you master the moves that turn a name in the pipeline into a closed deal — discovery that finds the truth, a clean capture that scores the deal, and a proposal that ties price to proof — run them on a real account, and prove, against a real rubric, that you can.

It’s Module 3 of the certifiable Sales track. It inherits the value narrative, qualification rubric, and battle-cards you built in Module 1 and runs on what the Module 2 outreach engine fills the calendar with — because a deal is only as good as the message it carries and the qualification behind it. Module 2 books the meeting; this module is where the meeting becomes a deal.

Prospecting fills the pipeline; running the deal is where it converts — and the difference is truth and proof. Discovery isn’t a feature tour; it’s where you surface the real problem, the budget, and who actually decides. A proposal isn’t a price list; it’s the buyer’s own words tied to the proof that makes the number feel inevitable. Master those and your pipeline stops being a list of names and starts being deals that actually close.

Every call has one job — or the deal stalls

The default failure is the feature tour: the same demo to every prospect, treating a first “yes, send me times” like buying intent, and torching the deal by showing instead of asking. A real deal is run — and the discipline is one rule applied without mercy: every call has exactly one job, and earns the next.

• Name the job before you dial. Before any agenda, each call gets a single job: discovery is for the truth, the demo is to show the fit you already qualified, the proposal review is to close. The moment a call tries to discover and demo and close, it does none of them. Decide the job first; the agenda is downstream of it.
• The asks escalate — no call assumes a yes the last one didn’t earn. The classic break is sending a proposal off a cold first call, or pitching pricing before you’ve surfaced a budget. A deal is an arc: surface the pain, confirm the fit, show the value, then earn the close. Each call is permission for the next ask, not a given.
• Map the deal stage to which call you’re on. Where they entered, what they already know, and what’s actually qualified decides which call this is — a discovery call is not a demo is not a close. Knowing which one you’re on is the difference between a conversation that moves the deal and one that stalls it by skipping a step the buyer needed.
• Claude preps and captures; you run the room. This is a hard line. Claude builds the question plan, the listening guide, and the structured capture — but reading the room, following the thread that opens up, and owning every commitment are yours. The playbook makes you prepared; it does not make the call for you, and nothing commercial commits without a person.

Discovery surfaces the truth, not a feature tour

Discovery is where deals are really won — and where reps either uncover the budget, the urgency, and the decision process, or wing a demo of the wrong thing weeks later. The craft of asking good questions is the easy part. The discipline most reps skip is that discovery is for truth, and a tailored capture turns that truth into a scored deal — not a warm feeling and a “circle back.”

• Lead with the problem; the money lands after the pain. Questions ordered to surface the real problem first, then its cost, urgency, budget authority, and decision process. Budget and urgency questions land far better once the buyer has talked themselves into the pain — ask “what’s this costing you today?” before “what’s the budget?” and you get a real number instead of a defensive one.
• A demo is not discovery. A feature tour answers questions nobody asked and teaches the buyer nothing about whether you fit. The hardest discipline on a good call is not solving out loud the moment you hear a problem you can fix — surface the whole truth first, qualify it, then show the fit on the next call. A strong demo of the wrong thing is the most expensive call in the pipeline.
• Map who decides — and find the champion. Who the economic buyer is, the real decision process, and whether there’s a champion with the buyer’s ear are what separate a real deal from a pleasant chat. If they can’t name who signs off, that’s the gap to close before a demo, not after the proposal stalls.
• Score the deal after the call, against the M1 rubric. Don’t ask Claude to “summarize the call” — you’ll get a play-by-play. Ask for the structured fields that decide what happens next: the problem in their words, the cost, the urgency, the budget and authority, the decision process, the competitor, the open questions — each filled or flagged [NEED] — then an honest A/B/C score against your icp.md rubric and the single biggest gap to close. That capture is what turns a conversation into a deal the whole team can read.

The commercial gate — consent’s analog: terms clear deal-desk, every number is fact-checked

A proposal is the strongest move you make and the single most likely thing to commit your team to something it didn’t mean to. The framing and the craft — mirror the buyer, tie value to proof — is the easy 60%. The 40% that gets teams burned is the part most courses never mention: the commercial gate. Three gates, and all three close before anything reaches the buyer.

• The number is never the rep’s to commit alone. Pricing, discounts, contract terms, and the expansion ask are drafted, then routed past deal-desk or a manager before the buyer sees them. Claude prices strictly from pricing.md and flags anything out-of-band — it must not invent a number, apply a discount beyond the approved bands, or commit a term. A quote that goes out ahead of sign-off is a commercial problem and a credibility problem at the same time.
• Every figure is fact-checked to the source. A proposal is near-contractual: a confident wrong price, an overclaimed ROI, or a scope line you can’t deliver becomes a problem you’re bound to. Every number traces to pricing.md or value-narrative.md, or it gets marked [NEED] and confirmed — never invented. Claude will happily turn a discovery aside into a hard ROI claim and state it with total confidence; that’s the limit you design around.
• Mind the legal and terms gate. A proposal can quietly carry a non-standard term, a delivery date you can’t hit, or a discount that sets a precedent across the pipeline. Route anything legal, regulated, or non-standard past the right reviewer, and put a named human on the price that goes out. Claude builds the apparatus — the sourced-number list, the who-signs-off list — but it doesn’t clear you. A person does.

The discipline here isn’t bureaucracy; it’s what lets you send something sharp while the deal is still warm, instead of a quote that blows up after signature because nobody owned the number.

Map the deal to the narrative — your foundation in motion

Here’s the connection that turns a competent deal into a sharp one: each step of the deal leans on a different piece of the foundation you built in Module 1. The deal isn’t separate from your value story — it’s your value story, qualification rubric, and battle-cards set in motion, one call at a time.

• Discovery confirms the pain your narrative was built to kill. The problem you surface should be the one value-narrative.md already names — the dread of the books, the VAT scramble, the decision made on month-stale numbers. If discovery surfaces a pain your narrative doesn’t address, that’s a fit signal to read, not a deal to force.
• The proposal ties price to the proof pillar. The number isn’t justified by a feature list; it’s anchored on the cost of doing nothing and carried by the pillar with real proof — Mizan’s “Trustworthy, not just fast,” every figure tracing to its source. Price against their cost of staying put, with a figure you can stand behind.
• The objection answers come from the battle-cards. When the proposal triggers “too expensive” or “I already have an accountant,” you don’t improvise — you reach for the M1 card: the honest reframe tied to proof, not a defensive dodge. The deal is the foundation under pressure.
• The test is simple: if a step can’t name the pillar, objection, or qualification signal it carries, it’s filler — and filler is where deals stall. A deal that maps cleanly to your foundation reads like one confident argument; one that doesn’t reads like a different rep on every call.

Bilingual deals — adapt, don’t translate

For a Gulf buyer the Arabic deal is a parallel motion, not a translation pass bolted on after the English ships. The rule is the same adapt, don’t translate you learned in the foundation — and a deal has its own edges most teams miss.

• The discovery localizes, not just the words. How directly you can ask about budget, how much the relationship comes before the spreadsheet, which question earns trust first — these differ. An Arabic discovery call that opens like a Western qualification script reads as foreign before the real conversation starts. Re-shape the call; don’t just re-language it.
• The proof and the lead objection differ. Who decides, what earns trust, and which pushback comes first are not the same across markets. The local proof — the FTA acceptance, the GCC business count — carries weight a generic global stat doesn’t, and “I already have an accountant” often comes first and hardest for the Gulf owner, because the family accountant is a real relationship, not a line item. Re-order the objections to what lands here.
• Register, authored from copy you’re proud of. The Arabic proposal is built from value-narrative-ar.md — lightly formal Modern Standard Arabic with Gulf-natural phrasing, not the stiff fus’ha of a bank letter, not slang — with numbers kept Western. The Arabic proposal is written, not converted.
• Same gate, both languages. The Arabic proposal clears the same deal-desk and legal review as the English. The failure mode is letting the second language skip sign-off because nobody on the approval chain reads it. Find someone who does — that’s the credibility line in this region, not a nicety.

Your assignment

Run a working deal for one product — your own (recommended: the output is a real deal your team moves) or the sample brand Mizan, the GCC small-business bookkeeping tool used throughout this track, selling to the sample prospect Najd Logistics. Open the folder with your context in Claude Desktop, approve each read in the “Ask permissions” prompt, drop your call transcript into the file pane, and work in the chat — no terminal needed.

Module 3 deliverable — running the deal

1. discovery-plan.md + a structured capture  (one new deal)
   - 6–8 open questions ordered problem → cost → urgency → budget →
     decision process, each with its strong-vs-weak tell
   - a one-glance listening guide (the signals to catch live)
   - a structured capture that fills every qualification field or flags
     it [NEED], and SCORES the deal A/B/C against the M1 rubric + the
     one gap to close next

2. post-call-recap.md  (the call captured into momentum)
   - the real objections — and the concern underneath each
   - what they committed to + the questions still open
   - a send-ready recap under 120 words with ONE dated next step
   - every number and commitment fact-checked, not remembered

3. proposal.md  (the deal, priced and pre-handled)
   - the buyer's problem in their own words, value tied to proof
   - scope + price calculated ONLY from pricing.md; anything out-of-band
     flagged for deal-desk, never invented
   - the 3 objections it will trigger, pre-handled inside the text
   - an Arabic version for bilingual teams — authored, not translated

Inherit the foundation: feed in value-narrative.md, icp.md (the
qualification rubric), battlecards.md, and pricing.md so the whole deal
stays on message, scores the same way twice, and prices from your real
rules instead of a guess.

The toolkit blocks below give you a fill-in template for each, so you’re filling in structure, not staring at a blank page.

How it’s graded — the rubric

This is the part the free playbooks don’t have, and the part that makes the credential mean something. Your three files are scored against five criteria. Each is meets / nearly / not yet — and “nearly” on any one is a revise, not a pass.

Deal rubric

1. Discovery surfaces truth     Questions lead with the problem; the call
                                surfaces real pain, budget, and who decides —
                                not a feature tour the buyer didn't ask for.
2. The capture scores the deal  Every qualification field is filled or flagged
                                [NEED]; the deal gets an honest A/B/C against
                                the M1 rubric, with the one gap to close named.
3. Proposal mirrors + proves    Opens in the buyer's own words; price is anchored
                                on the cost of doing nothing and tied to a proof
                                pillar — not a feature list with a number stapled on.
4. Commercial gate held         Pricing/terms cleared deal-desk before the buyer;
                                every number traces to pricing.md or the narrative;
                                gaps are [NEED], not invented; out-of-band flagged.
5. Arabic is adapted            The Arabic proposal is authored, not translated:
                                proof localized, lead objection re-ordered,
                                register right, same sign-off gate. (Bilingual.)

The discipline is deliberately what a senior sales leader would demand: a discovery call that’s a feature tour, a proposal with a number nobody can source, or an Arabic proposal that reads translated fails in front of the buyer — so it has to be caught here, against the rubric, not live on a call or after signature.

The bar, shown — a worked model answer (Mizan)

You don’t have to guess what “meets” looks like. Here are passing excerpts for the sample deal — yours doesn’t need to match these, it needs to clear the same bar. Mizan’s English copy stays English by design.

discovery-capture.md — Najd Logistics (excerpt)

Account: Najd Logistics — 30 staff, 3 branches, VAT-registered.
         Owner does the books in Excel on weekends.

The real problem (their words)
  "By the time I see last month's numbers, the month's already gone. And
   every VAT quarter I lose a weekend reconciling three branches by hand."

Cost / impact
  ~2 weekends a quarter on VAT; one late filing last year (AED 1,000 penalty).
  Considering a part-time bookkeeper (~AED 1,500/mo).

Urgency / the trigger ("why now")
  Opening a 4th branch in Q3 — Excel won't stretch to four.

Budget & authority
  Owner is the economic buyer; signs himself. Comfortable "around AED 500/mo
  if it saves the weekends." Hasn't seen a number yet.         ← [NEED] confirm

Decision process
  Owner decides; wants his accountant to "bless it" first.  Champion: owner.

Deal score (against icp.md rubric)
  6/6 fit signals, 0 disqualifiers, clear trigger, named buyer → A / STRONG.
  Biggest gap to close next: get the accountant on a call so "bless it"
  can't quietly stall the proposal.

post-call-recap.md — Najd Logistics (excerpt)

Real objections (and the concern underneath)
  "Let me run it past my accountant."  → real concern: does this replace him /
                                          threaten the relationship? (it doesn't)
  "Can it handle three branches in one view?" → needs proof of consolidation.

Committed to
  Owner will set up a 20-min call with his accountant this week.

Open question we can answer in writing
  Yes — consolidated multi-branch books are the Multi-branch tier.

Recap email (send now, < 120 words)
  "Thanks, [owner] — good call. You flagged that last month's numbers arrive
   too late, and that VAT quarters eat a weekend across the three branches.
   Mizan closes all three in one view and prepares the VAT return ready to
   file — every figure tracing to its source. Keep your accountant; he moves
   to advice, not data entry. As agreed, you'll set up a short call with him
   this week, and I'll send a proposal right after so he can bless the
   numbers. Talk soon."

Next step: accountant call this week → proposal to follow.   (dated, specific)

proposal.md — Najd Logistics (excerpt)

Their problem, their words
  "By the time I see last month's numbers, the month's already gone — and VAT
   across three branches eats a weekend every quarter."

The outcome, tied to proof
  All three branches closed in one view; a VAT return ready to file. Every
  figure traces to its source document — the proof behind "trustworthy, not
  just fast": 9,000+ GCC businesses; 2025 VAT exports the FTA accepted
  without amendment.

Scope & price (from pricing.md — Multi-branch tier)
  Multi-branch:            AED 499/mo
  + 3 branches × AED 79:   AED 237/mo
  ───────────────────────────────────
  List:                    AED 736/mo   (annual = 2 months free)
  Consolidated books across all branches + FTA auto-filing.

Price tied to the cost of doing nothing
  One late VAT filing last year cost AED 1,000+. A part-time bookkeeper runs
  ~AED 1,500/mo — more than twice this, for data entry Mizan already does.

Objections, pre-handled in the text (from battlecards.md)
  "I already have an accountant" → keep him; he moves to advice, not entry.
  "Why now" → the 4th branch opens in Q3; Excel won't stretch to four.

Commercial-gate note (what a person cleared)
  • Price is straight off pricing.md list — no discount → no deal-desk needed.
  • Owner floated "annual at 3 months free" — that's BEYOND the standard
    2-months-free band → flagged for deal-desk, NOT promised in the draft.
  • The "saves ~AED 1,500/mo vs a bookkeeper" line is the owner's own figure
    from discovery, not yet confirmed → marked [NEED] until verified.
  • A named human (the rep's manager) signs the number before it sends.

Arabic note (authored, not translated)
  Built from value-narrative-ar.md. Lead objection re-ordered: «عندي محاسب»
  opens the objections, warmest, because the family-accountant relationship
  is real for the Gulf owner. Proof stays local (FTA, the GCC count).
  Numbers Western. Same deal-desk + manager sign-off, cleared by an Arabic
  reader on the approval chain.

What you keep — toolkit additions

These extend the Foundation toolkit — copy them into the same folder your team works in, replace the Mizan placeholders with your own, and every deal starts from structure instead of a blank page.

The discovery-plan template — the problem-first plan and the scored capture, made fill-in-the-blank.

# discovery-plan.md — [Product] · [account]

Goal of THIS call (one sentence): [discovery — surface the truth, not demo]

Questions (problem first, then the money and the process):
1. [open question — the real problem]   strong tell: [...]  weak tell: [...]
2. [its cost / impact]                  strong tell: [...]  weak tell: [...]
3. [urgency / the trigger — "why now"]  ...
4. [budget + who has authority]         ...
5. [decision process + the champion]    ...
(+ 2 follow-up probes for the answers you most need to dig into.)

Listen for (keep in a corner of the screen):
[real pain · a number · a deadline · who decides · a competitor]

── after the call: the capture ──
Problem (their words) · cost/impact · urgency · budget + authority ·
decision process · competitor · open questions
Each field filled, or marked [NEED] — never guessed.
Score against the rubric: A / B / C + a one-line reason + the one gap to close.

The proposal template + the commercial-gate checklist — the part that turns a draft into something you can actually send.

# proposal.md — [Product] · [account]

Their problem (their words):  [echo discovery — proves you listened]
The outcome, tied to proof:   [value + the proof pillar that matters to them]

Scope:  [what's included — only what discovery confirmed]
Price:  [tier + add-ons, calculated FROM pricing.md — show the math]
        [annual terms only within the approved band]
Price vs. cost of doing nothing: [the penalty / the wasted spend they named]

Objections, pre-handled (from battlecards.md):
- "[likely objection]" → [the reframe, in the text]
- "[likely objection]" → [the reframe, in the text]

Commercial gate (every box before it reaches the buyer):
[ ] Every number traces to pricing.md or value-narrative.md — or it's [NEED]
[ ] Any discount or term beyond the approved band → flagged for deal-desk
[ ] Nothing legal / regulated / non-standard left un-reviewed
[ ] A named human owns the price that goes out
→ Any box open? It does not reach the buyer yet.

The saved prompts — paste them in the chat with the right files open. (On the optional Power Track these become slash commands; you need none of that to use them today.)

Build a discovery plan tailored to this account
  "Read value-narrative.md, icp.md, and what we know about [account]. Build a
   discovery plan: 6–8 open questions ordered to surface the real problem
   first, then its cost, urgency, budget authority, and decision process. For
   each, note what a strong vs weak answer tells me, plus 2 follow-up probes.
   One page I can glance at on the call. Don't script it — I'll run the room."

Turn a transcript into a recap — every claim flagged for fact-check
  "Read [transcript].txt. Pull four lists: every objection (and the real
   concern underneath each), what they committed to, the open questions, and
   the buying signals. Then draft a recap under 120 words with ONE dated next
   step. Flag every number, date, and commitment with [NEED] for me to verify
   against my own memory — don't promise back anything they didn't agree to."

Draft a proposal that prices only from pricing.md
  "Read the discovery capture, value-narrative.md, battlecards.md, and
   pricing.md. Draft a proposal: the buyer's problem in their words, value
   tied to proof, scope + price calculated ONLY from pricing.md (show the
   math), and the 3 objections it'll trigger pre-handled in the text. Flag
   anything needing a discount beyond the approved band or a number you can't
   ground, for deal-desk. Do NOT invent a price or commit a term."

Adapt the deal for the Gulf — don't translate it
  "Read value-narrative-ar.md, battlecards.md, and pricing.md. Adapt this
   proposal into Arabic for a Gulf buyer: match the Arabic register, localize
   the proof, re-order the objections if 'I already have an accountant' leads
   here, keep numbers Western. This is an adaptation, not a translation — and
   it clears the same deal-desk sign-off as the English."

What you’ve proven — and what’s next

Clear the rubric and you’ve proven something the free playbooks can’t certify: you can take a name in the pipeline and run it into a deal — discovery that finds the truth, a capture that scores it honestly, and a proposal that ties price to proof and clears the commercial gate. That’s the Deal stage of “Certified Sales with Claude.”

From here the track turns a single deal into a repeatable motion, each module assessed the same way:

• Module 4 — the big play: running the multi-stakeholder, multi-threaded deal — the larger account where one champion isn’t enough and the close depends on orchestrating several people without losing the thread.
• Module 5 — measure & forecast, then the capstone — one account taken from cold to won-and-growing, end to end, graded into the certificate.

Because this module lives or dies on the commercial gate, pair it with the operating guide — the data, deal-desk, and sign-off layer that belongs underneath every proposal you send and every number you quote. It’s the safety net that makes “Claude drafts; a person decides” a process, not a hope.