Case Studies: Pitch Deck, Exec Readout & Sales Deck

This final lesson walks through three complete deck workflows end to end, using everything from the previous nine lessons. You will see how the techniques compound across a real deck and how to recognize when to use which tool at which stage.

Each case is a fictional but realistic scenario you can mirror for your own work.

What You'll Learn

• A complete pitch deck workflow from blank page to ready-to-send
• A complete exec readout workflow for a board meeting
• A complete enterprise sales deck workflow
• Time benchmarks and tool-by-tool sequencing for each

Case Study 1: Series A Pitch Deck — VendorFlow

Scenario

VendorFlow is a fictional pre-Series-A startup. ARR $1.4M, growing 12 percent month over month. Founder is preparing for a 30-minute Series A meeting with a top-tier US fund. She has 4 hours total before the meeting.

Hour 1: Foundation — Audience, Arc, Brand Brief

Tools: Claude (long context for full arc)

Step 1: Paste the VC fund's most recent investor letter into Claude.

Below is the latest investor letter from [Fund]. Generate an audience brief I will use for every prompt in this pitch deck. Capture: role, decision they are about to make, what is on their mind, vocabulary, and what annoys them.

Step 2: Generate the Investor Arc.

Audience brief: [paste]

Company in one paragraph: VendorFlow replaces the 9-week paperwork phase of vendor onboarding with a 3-day self-serve flow. We have $1.4M ARR, 12% MoM growth, 9 mid-market customers including [two named anchor accounts]. Founders are [1-line credentials each].

Generate a Problem / Solution / Why Now / Why You arc. For each: 1-sentence claim, 3 supporting points, most likely VC pushback. Then map to a 12-slide deck outline with action titles. Then run a critique pass and re-output the refined version.

Step 3: Paste the brand brief and lock the visual style.

Time spent: 50 minutes. End of hour 1: clean arc, 12 action titles, brand brief, audience brief — all stored in one Claude conversation.

Hour 2: First Draft — Slide Bodies and Numbers

Tools: Claude (continuing the conversation), Perplexity (market data)

Step 4: For each of the 12 slides, request slide content.

For slide [N], action title "[title]", generate: (1) central visual concept, (2) 3 supporting bullets, (3) data source for any number cited, (4) the 30-second talk track.

Batch 5 slides at a time. Roughly 5 minutes per batch.

Step 5: Cross-check market sizing in Perplexity.

Confirm with cited sources: there are approximately X mid-market US companies in the $100M-$2B revenue range. List at least 3 sources.

Step 6: Run the simulated VC partner-meeting prompt.

Below is the full deck content. Simulate a partner meeting with 3 partners (generalist, finance operator, vertical specialist). Each asks 3 hard questions. Give a fund/pass verdict and the 3 things that would change the verdict.

Time spent: 60 minutes. End of hour 2: complete slide content, sourced numbers, identified the 3 weakest spots in the deck.

Hour 3: Build the Actual Deck

Tools: Gamma for first draft, PowerPoint for final polish

Step 7: In Gamma, paste the full outline:

Generate a 12-slide pitch deck from the outline below. Use my brand kit. Action titles exactly as written. One main visual per slide. 16:9. [paste outline]

Gamma produces a near-finished deck in about 60 seconds.

Step 8: Export to PowerPoint. Drop into VendorFlow's master template.

Step 9: Replace generic Gamma illustrations with brand-consistent images. Use the style-locked image prompt from Lesson 7.

Time spent: 60 minutes. End of hour 3: visually consistent deck in the master template.

Hour 4: Speaker Notes, Q&A, Dress Rehearsal

Tools: Claude

Step 10: Generate time-budgeted speaker notes (30-minute meeting, ~18 minutes for the deck, ~12 for Q&A).

Step 11: Run the hostile Q&A simulator (20 questions: 5 easy, 10 hard, 5 curveball).

Step 12: Drill the 5 weakest answers with the critique prompt.

Step 13: Run the dress rehearsal prompt to produce a full walkthrough plus a 90-second elevator-pitch fallback.

Time spent: 60 minutes. End of hour 4: deck, talk tracks, Q&A binder, dress-rehearsal artifact.

Total: 4 hours. Result: A deck and prep package that, done manually, would have taken 20-25 hours.

Case Study 2: Board Readout — NorthStar Health

Scenario

NorthStar Health is a fictional Series B healthtech company. The CFO is preparing a 20-minute Q2 board readout. The audience: 5-person board including 2 investor partners, 2 independent directors, the CEO. Material: financial results, AI risk discussion, H2 plan ask.

Hour 1: Arc and Recommendation

Tools: ChatGPT

Use the SCQA Board Arc prompt:

Situation: H1 closed at 92% of plan, ARR $28M, gross retention 91%, NDR 117%. Complication: AI-coded competitors have entered the market and price-discounted aggressively. We saw 2 churned accounts in May where a competitor was 40% cheaper. Recommendation: Accelerate our AI-native rebuild of the workflow engine and reduce headcount growth in H2 to fund it.

Generate the SCQA executive summary, then the 7-section deck outline with action titles. Then list the 3 most likely board objections and how to handle each.

End of hour 1: clean recommendation, deck outline, anticipated objections.

Hour 2: Build the Deck

Tools: PowerPoint (using NorthStar's template) with Copilot

Step 1: Open the NorthStar board master template in PowerPoint.

Step 2: Use Copilot to convert the outline into slides:

Build slides from the outline below. Use the existing master template's layouts, colors, and fonts. One idea per slide. Action titles as written. Insert placeholder charts where the outline mentions numbers; I will replace them. [paste outline]

Step 3: Replace placeholder charts with real financial charts pulled from the company's BI tool. Use Copilot for chart action titles ("Revenue grew 18% vs Q1, driven by mid-market segment").

Step 4: Insert the AI risk discussion slide. Use the design-rationale prompt:

Audit the AI risk slide. Why this chart type? Why this color? Why this layout? What alternative was rejected? How would I defend each choice in the board meeting?

Hour 3: Speaker Notes, Anticipated Q&A, Rehearsal

Step 5: Generate 30-60-90 framework speaker notes for a 20-minute slot.

Step 6: Hostile Q&A simulation — board edition. The curveball questions for a board are different from a VC. AI tends to surface:

• "What is the dollar impact if our second-largest customer churns?"
• "Why are we increasing headcount in engineering when the gross margin trend is flat?"
• "If AI commoditizes our workflow, what is the irreversible thing we are betting on?"

Step 7: Pre-circulate the deck 48 hours before the meeting (board norm). Note: AI does not absolve you from following good board hygiene.

Total: 3 hours. Result: a board-ready deck plus prep, vs the typical 8-12 hours.

Case Study 3: Enterprise Sales Deck — TaskLoop

Scenario

TaskLoop is a fictional B2B SaaS product. A senior AE is preparing a custom 15-slide deck for a meeting with a Fortune 500 CFO who has a procurement-driven 90-day evaluation cycle. The deck must be account-specific.

Hour 1: Account Research + Sales Arc

Tools: Perplexity, ChatGPT

Step 1: Research the account in Perplexity.

Research [Account Name]. Return: (a) their last 10-K key themes, (b) recent earnings call topics, (c) recent press releases on operational priorities, (d) any references to AI, automation, or process efficiency initiatives. Cite sources.

Step 2: Feed the research into ChatGPT with the Sales Arc prompt.

Audience brief: CFO at [Account]. Recent operational priorities from their earnings call: [paste].

Generate a Status Quo / Tension / Resolution / Proof narrative for TaskLoop. Open the deck with this CFO's specific situation — quote their own priorities back to them. Then map to a 15-slide outline with action titles.

This is the highest-impact prompt in the sales workflow: opening with the buyer's own words, sourced from public artifacts, signals "we did our homework" before slide 1.

Hour 2: Slide Bodies and Custom Visuals

Step 3: For each slide, request slide content with the audience brief loaded. Pay special attention to the "Proof" slides — they need account-relevant customer stories.

Step 4: Build a custom ROI calculator slide. Prompt:

Build a custom ROI estimate for [Account]. Inputs: their estimated headcount of [N] in the relevant function, current process cost assumption of [range], our typical 40-60% efficiency gain. Output: a 3-line ROI estimate with a methodology footnote. Include a "this is illustrative, methodology to be validated jointly" disclaimer.

The disclaimer is important. CFOs hate vendor ROI math that overpromises; the disclaimer plus a willingness to validate jointly turns the slide from a sales pitch into a discussion starter.

Hour 3: Q&A and Talk Tracks

Step 5: Hostile Q&A simulator — CFO edition. Curveball questions AI surfaces include:

• "What is the smallest commitment you would accept to start?"
• "Walk me through your worst customer outcome. Why did it go wrong?"
• "If we delay this by two quarters, what specifically goes wrong for us?"

Step 6: Talk tracks. Convert each slide's speaker notes to: 1 must-say sentence + 3 talking points + 1 transition. CFO meetings often have 4 to 5 interruptions in the first 10 minutes.

Total: 3 hours for a fully customized account-specific 15-slide deck. Without AI, this is typically a half-day per account; AEs simply skip the customization and use the generic deck, which is why most sales decks underperform.

What These Three Cases Have in Common

Pull back and a pattern emerges across all three workflows:

1. Audience first, slides last. Every case opens with audience research and brief construction, not with slide design.
2. Arc before content. The full narrative arc is locked before any slide content is written.
3. AI for thinking, generation tools for production. Claude/ChatGPT do the structural and narrative work; Gamma, Copilot, and PowerPoint do the production work.
4. Always simulate hard questions. The hostile Q&A prompt is in every case. It is the single most valuable AI deck habit.
5. Brand and design are enforced in prompts, not after the fact. A brand brief at the top of every prompt produces 80% on-brand output. Fixing inconsistencies after generation is far slower.

Three to four hours, end to end, for a deck that used to take a full day or more. That is the realistic ROI of mastering this workflow.

A Final Habit: Build Your Prompt Library

Every prompt in this course is reusable. The single highest-leverage thing you can do after finishing this course is to save these prompts in a personal library — a Notion page, a Google Doc, or a custom GPT.

A simple library structure that works:

• Audience briefs (one per persona type)
• Arc prompts (Investor, Board, Sales)
• Slide content prompts (Problem, Solution, Why Now, Market, Competition, Traction, Business Model, Financials, Ask, Team, Appendix)
• Visual prompts (hero image, persona portrait, diagram, style lock)
• Delivery prompts (time-budgeted speaker notes, talking points, Q&A simulator, drill, dress rehearsal)
• Audit prompts (coherence audit, deck audit, design rationale)

The first deck you build with this library takes 4 hours. The fifth takes 90 minutes. The twentieth takes 45 minutes — and by then you are noticeably better at the underlying craft, not just faster.

Key Takeaways

• A complete Series A pitch deck takes about 4 hours with this workflow vs 20-25 manually
• A board readout takes about 3 hours vs 8-12 manually
• A fully customized enterprise sales deck takes about 3 hours per account vs a half-day
• The five universal habits: audience first, arc before content, AI for thinking and Gamma/Copilot for production, simulate hard questions, enforce brand in prompts
• Build a personal prompt library after this course — by deck five you are at ~90 minutes per deck