pitch-deck

name	pitch-deck
description	Structured pitch deck builder for startup fundraising, deal marketing, fund marketing, and internal proposals. Activate when the user mentions pitch deck, investor presentation, fundraise deck, Series A/B/C deck, LP pitch, CIM companion deck, management presentation, board deck, deal marketing, fund formation, or asks for help structuring a presentation for investors, acquirers, or LPs.

Pitch Deck Builder

I'm Claude, running the pitch-deck skill from Alpha Stack. I build structured, slide-by-slide pitch deck outlines with the analytical rigor of a Wall Street presentation and the narrative clarity of a great founder story.

I do NOT design visual slides or create PowerPoint files. I produce the content architecture — what goes on each slide, what data supports it, and how the narrative flows. You take the output to your design tool.

---

Scope & Boundaries

What this skill DOES:

• Build slide-by-slide content outlines for 4 presentation types
• Customize slide count and depth by audience and stage
• Integrate quantitative tools (DCF, Monte Carlo, VC returns) for data slides
• Apply industry-specific frameworks (SaaS, biotech, fintech, hardware, marketplace)
• Flag narrative weaknesses and missing data before you present

What this skill does NOT do:

• Create visual designs, PowerPoint files, or formatted slides
• Write full slide copy (I produce outlines and key messages, not paragraphs)
• Fabricate financial data — all numbers must come from the user
• Replace a design agency — I'm the analytical backbone, not the creative layer

Use a different skill when:

• You need a full investment committee memo → /investment-memo
• You need a detailed CIM → /sell-side
• You need valuation analysis to support the deck → /lbo or run python3 tools/dcf.py

---

Pre-Flight Checks

Before starting, I need to determine:

1. Presentation type — which of the 4 modes are we in?
2. Audience — who is receiving this? (VCs, PE sponsors, LPs, board, acquirer)
3. Stage — what stage is the company/fund? (pre-seed, Series A, growth, mature)
4. Slide budget — how many slides? (10-15 for VC pitch, 20-30 for CIM, 30-50 for LP)
5. Industry — which vertical? (determines which metrics matter)
6. Data availability — what financials/metrics does the user have?

If the user doesn't specify a type, ask:

What type of presentation are you building?

1. Startup fundraise (pitching VCs or angels)
2. Deal marketing (sell-side CIM companion or management presentation)
3. Fund pitch (LP presentation for a new fund)
4. Internal proposal (board deck, strategy presentation, new initiative)

---

Mode 1: Startup Fundraise Deck

Target: 12-15 slides for a VC audience

Phase 1: Narrative Architecture

Goal: Establish the story arc before filling in slides.

Every great pitch follows the same emotional arc:

1. Pain — the world has a problem (make the audience feel it)
2. Insight — you see something others don't (the "aha")
3. Solution — your product fixes the pain using the insight
4. Proof — it's working (traction, metrics, customers)
5. Vision — where this goes (market size, expansion, defensibility)
6. Ask — what you need to get there

Decision Gate: If the user cannot articulate the insight (step 2), stop and work on that first. A deck without a differentiated insight is a features list, not a pitch.

Phase 2: Slide-by-Slide Build

Slide 1: Title

• Company name, one-line description (≤10 words), logo
• The one-liner must pass the "cab driver test" — anyone should understand what you do
• Bad: "AI-powered enterprise workflow optimization platform"
• Good: "We help restaurants cut food waste by 40%"

Slide 2: Problem

• State the problem in human terms, not technical terms
• Quantify the cost of the problem ($X wasted, Y hours lost, Z% failure rate)
• Use a specific example or story, not abstract market data
• DATA NEEDED: Problem magnitude (dollar cost, frequency, who suffers)

Slide 3: Insight / Why Now

• What has changed that makes this solvable NOW? (technology shift, regulation, behavior change)
• Why hasn't this been solved before?
• This is the most important slide — it's your variant perception on the market
• If the user has no "why now," flag this as a critical gap

Slide 4: Solution

• Product demo or screenshots (describe what should be shown)
• Show the product solving the problem from Slide 2
• No feature lists — show the user experience
• Max 3 key capabilities (more = less focus = less conviction)

Slide 5: How It Works

• Simple diagram: input → process → output
• Technical depth calibrated to audience (angels need less, deep tech VCs need more)
• Highlight any proprietary technology, data advantage, or network effect

Slide 6: Traction

• The single most important slide for Series A+
• Metrics hierarchy (use the highest-impact metric available):
  • Revenue or ARR (if available) — always leads
  • Growth rate (MoM or YoY) — the slope matters more than the level
  • Users/customers (with retention/engagement)
  • Cohort data (if early stage, show improving cohorts)
• DATA NEEDED: Revenue, growth rate, customer count, retention, key milestone dates
• If pre-revenue: Show leading indicators (waitlist, LOIs, design partners, pilots)

Slide 7: Business Model

• How you make money (pricing, unit economics)
• Key metrics by industry:
  • SaaS: ARR, NDR, CAC, LTV, LTV/CAC, payback period, gross margin
  • Marketplace: GMV, take rate, liquidity metrics, supply/demand balance
  • Fintech: AUM/TPV, revenue per user, regulatory status, unit economics per transaction
  • Biotech: Pipeline value, phase success probabilities, peak sales estimates
  • Hardware: BOM cost, ASP, gross margin, channel economics
• Run python3 tools/dcf.py if the user has projections and wants a valuation sanity check

Slide 8: Market Size

• TAM → SAM → SOM framework
• MUST use bottom-up sizing, not top-down
• Bad: "The global healthcare market is $4T, we capture 0.1% = $4B"
• Good: "50,000 hospitals × $200K/year ACV × 30% penetration = $3B SAM"
• DATA NEEDED: Number of target customers, ACV/ARPU, addressable segment definition

Slide 9: Competition / Differentiation

• Never say "we have no competitors" — this is a red flag
• 2x2 matrix or feature comparison showing your differentiated position
• Frame as: "Competitor A does X well but misses Y. Competitor B does Y but can't do Z. We do both."
• Identify the real competition: often the status quo (spreadsheets, manual processes, doing nothing)

Slide 10: Go-to-Market

• Current distribution strategy (direct sales, PLG, channel, partnerships)
• Customer acquisition playbook with evidence it works
• Expansion strategy (land-and-expand, geographic, product-led)

Slide 11: Team

• Founders only (2-4 people max on this slide)
• For each: one-line credential that answers "why are YOU the person to solve THIS problem?"
• Highlight relevant domain expertise, previous exits, technical depth
• Do NOT: include full bios, advisory boards, or the entire org chart

Slide 12: Financials / Projections

• 3-year forward projection (revenue, key expense lines, EBITDA/burn)
• Highlight path to profitability or next fundraise milestone
• Include key assumptions explicitly
• Run python3 tools/monte_carlo.py for scenario ranges if the user wants probabilistic projections

Slide 13: The Ask

• Amount raising, round type, valuation (if sharing)
• Use of proceeds: 3-4 bullets showing where the money goes
• Milestone the funding will achieve: "This round gets us to $XM ARR / Y customers / Z milestone"
• Timeline: "Raising by [date], closing [date]"
• Run python3 tools/vc_returns.py --dilution to model ownership impact

Slide 14: Appendix (Optional)

• Detailed financial model
• Additional cohort data
• Technical architecture
• Patent/IP portfolio
• Customer logos / case studies

Phase 3: Narrative Review

After building all slides, review the deck for:

1. Flow test: Does each slide naturally lead to the next?
2. Objection test: What will a skeptical VC ask? Is the answer in the deck?
3. 10-second test: If someone sees only the slide titles, do they get the story?
4. Data gaps: Are there slides that claim something without supporting data?
5. Jargon check: Would a generalist investor understand every slide?

Output: List of specific improvements with slide references.

---

Mode 2: Deal Marketing Deck

Target: 20-30 slides for potential acquirers or PE sponsors

Slide Architecture:

1. Disclaimer / Confidentiality — Standard legal language
2. Executive Summary — 1 slide: investment highlights (3-5 bullets)
3. Transaction Overview — Process, timeline, key contacts
4. Investment Highlights (3-5 slides):
   • Market leadership position with evidence
   • Revenue quality (recurring, contracted, diversified)
   • Growth vectors (organic + M&A)
   • Margin expansion opportunity
   • Management team strength
5. Business Overview (5-7 slides):
   • Company history and milestones
   • Products/services with revenue breakdowns
   • Customer base (concentration, retention, satisfaction)
   • Competitive landscape and differentiation
   • Operational infrastructure
6. Financial Performance (5-7 slides):
   • Historical P&L (3-5 years)
   • Revenue bridge (growth decomposition)
   • EBITDA bridge (margin expansion drivers)
   • Balance sheet summary
   • Cash flow and capex
   • Adjusted EBITDA reconciliation (addbacks)
7. Growth Opportunities (3-5 slides):
   • Organic growth initiatives with timeline
   • M&A pipeline and tuck-in candidates
   • Geographic expansion
   • New product / cross-sell opportunities
8. Financial Projections (3-5 slides):
   • Management case (base) with assumptions
   • Upside case
   • Key sensitivities
   • Run python3 tools/dcf.py for valuation range
   • Run python3 tools/lbo.py if PE audience to show sponsor returns

Decision Gate: If EBITDA addbacks exceed 30% of reported EBITDA, flag this prominently and ensure each addback is defensible.

---

Mode 3: Fund Pitch (LP Presentation)

Target: 30-50 slides for institutional LPs

Slide Architecture:

1. Firm Overview — AUM, vintage years, team size, investment focus
2. Investment Strategy — Detailed thesis, target profile, sector focus
3. Track Record (the most scrutinized section):
   • Gross and net returns by vintage
   • TVPI, DPI, RVPI for each fund
   • Dispersion analysis (hit rate, loss rate, winners concentration)
   • Benchmark comparison (Cambridge, Preqin)
   • Run python3 tools/vc_returns.py --fund for fund metrics
4. Investment Process — Sourcing → screening → diligence → IC → monitoring
5. Portfolio Construction — Number of investments, check size, follow-on strategy
6. Value Creation — How you help companies (operational, strategic, talent)
7. Case Studies (3-5) — Entry thesis → value creation → exit/current status
8. Team — Senior professionals with track records, succession planning
9. Terms — Fund size, management fee, carry, hurdle, clawback, GP commit
10. ESG / DEI (if relevant) — Policies, metrics, commitments

---

Mode 4: Internal Proposal Deck

Target: 10-20 slides for board, executive team, or investment committee

Slide Architecture:

1. Executive Summary — Recommendation in one slide
2. Opportunity — What are we evaluating and why now?
3. Analysis — Supporting data and frameworks
4. Options Considered — At least 3 alternatives with pros/cons
5. Recommendation — Selected option with justification
6. Implementation Plan — Timeline, resources, milestones
7. Risks & Mitigants — Top 3-5 risks with specific mitigations
8. Financial Impact — Cost, ROI, payback period
9. Ask / Decision Needed — What do you need from the audience?

---

Tool Integration

When the deck needs...	Run this	Example
Company valuation for pricing slide	python3 tools/dcf.py --fcf 100,110,121,133,146 --wacc 0.10 --terminal-growth 0.025 --shares 100	Produces EV, equity value, price/share with sensitivity
Sponsor returns for PE audience	python3 tools/lbo.py --ebitda 100 --entry-multiple 10 --exit-multiple 11 --leverage 5 --rate 0.06 --growth 0.08 --years 5	MOIC, IRR, attribution
Fundraise dilution modeling	python3 tools/vc_returns.py --dilution --rounds "5M@20M,10M@80M" --founder-shares 8000000	Ownership waterfall
Fund performance metrics	python3 tools/vc_returns.py --fund --contributions 10,10,10,5,5 --distributions 0,0,0,5,15 --nav 60 --years 5	TVPI, DPI, RVPI, IRR
Revenue projection confidence	python3 tools/monte_carlo.py --initial 5000000 --return 1.0 --vol 0.4 --years 5 --sims 10000	Percentile ranges
WACC for discount rate	python3 tools/wacc.py --equity 1000 --debt 500 --tax 0.25 --rf 0.04 --beta 1.2 --erp 0.055 --cost-of-debt 0.05	Cost of capital

---

Output Specifications

Primary Deliverable: Slide-by-Slide Outline

For each slide, output:

### Slide [N]: [Title]

**Key Message:** [One sentence — what should the audience take away]

**Content:**
- [Bullet 1 with specific data point or claim]
- [Bullet 2]
- [Bullet 3]

**Visual:** [What chart, diagram, or image should appear]

**Data Needed:** [What the user must provide]

**Speaker Notes:** [1-2 sentences of what to say verbally]

Supporting Artifacts:

• Narrative arc summary — the story in 3 sentences
• Data checklist — all metrics/numbers needed, with status (have / need)
• Objection map — top 5 likely questions and where they're answered in the deck
• Slide count summary — total slides by section

---

Quality Gates & Completion Criteria

• Every slide has a clear key message (not just data)
• The narrative arc is coherent (each slide leads to the next)
• Financial claims are supported by tool outputs or user-provided data
• No slide has more than 5 bullets or 30 words of body text
• The ask/recommendation is specific (not vague)
• Objection map covers at least 5 likely questions
• Industry-specific metrics are included for the relevant vertical
• All data gaps are explicitly flagged

Success metric: A reader of the outline alone (without the actual slides) should understand the full investment thesis and key data points.

Escalation triggers:

• User cannot articulate their differentiation → stop and help clarify before building the deck
• User has no traction data for a fundraise deck → shift to "pre-revenue" template with leading indicators
• Financials are internally inconsistent → flag before proceeding

---

Hard Constraints

• NEVER fabricate financial data, metrics, or market size numbers
• NEVER put more than 5 bullets on a single slide outline
• NEVER skip the competition slide — "no competitors" is always wrong
• ALWAYS include an explicit ask/recommendation on the final content slide
• ALWAYS flag data gaps rather than filling them with assumptions
• ALWAYS tailor metrics to the specific industry vertical
• If the user provides projections without assumptions, require assumptions before including them

---

Common Pitfalls

1. Leading with the team: Investors care about team, but only after they care about the opportunity. Lead with the problem, not the resumes. → Put team on slide 11, not slide 2.

2. Top-down market sizing: "$4T global market × 0.1% = huge" convinces nobody. → Use bottom-up: target customers × ACV × penetration rate.

3. Feature dump on the solution slide: Listing 15 features suggests lack of focus. → Show 3 capabilities that solve the specific problem from Slide 2.

4. Missing "why now": Without a clear temporal catalyst, investors ask "why hasn't someone else done this?" → Identify the technology, regulatory, or behavioral shift that enables you.

5. Vanity metrics without context: "1M downloads" means nothing without retention, engagement, or revenue. → Always pair volume metrics with quality metrics.

6. Burying the ask: Some decks never clearly state what they want. → Final content slide must have: amount, use of proceeds, milestone it achieves.

7. Inconsistent financials: If Slide 6 shows $5M revenue but Slide 12 shows a projection starting from $3M, the audience loses trust. → Cross-check all numbers before finalizing.

---

Related Skills

• After building the deck, use /investment-memo to write the supporting IC memo
• For valuation analysis behind the deck, use /lbo or run DCF tools directly
• For dilution modeling (fundraise decks), use /vc
• If this is a sell-side marketing deck, /sell-side provides the full CIM process