pitch-deck

Pitch Deck & Financial Model Guide

A pitch deck sells one belief: that this team can turn this opportunity into an outsized outcome. Every slide either advances that belief or wastes a slide. The deck tells the story; the model proves it isn't fiction. Build both, and make them agree.

Core principle: a deck is a narrative, not a data dump. One idea per slide, the idea in the headline, the slide as evidence for it. If a slide doesn't move the story forward, cut it.

The canonical deck arc (10–12 slides)

Order is a story: a problem worth solving → a solution that works → why it wins → why now → how big → how it makes money → that it's already working → who's building it → the ask. Each slide has a job — to prove something, not just state it.

#	Slide	What it must prove
1	Cover / one-liner	What you do, in plain language a stranger gets in 5 seconds
2	Problem	A real, painful, expensive problem — for a specific someone
3	Solution	Your product solves that problem; show it, don't describe it
4	Why now	A shift (tech, regulation, behavior, cost) that makes this possible/urgent today
5	Market (TAM/SAM/SOM)	The prize is big enough to matter, and built bottom-up not hand-waved
6	Product	It's real and differentiated — screenshots, demo, the "magic moment"
7	Business model	How you make money, what a customer is worth, that the unit math works
8	Traction	Evidence it's working — the single most persuasive slide if you have it
9	Go-to-market	A repeatable, affordable path to customers (not "we'll go viral")
10	Competition	You understand the landscape and have a defensible edge
11	Team	Why this team wins this — relevant, credible, complementary
12	The ask	How much, for what milestones, and what it buys the investor

Optional add-ons: vision/roadmap, financial summary, use-of-funds breakdown. Keep the core to ~12; park detail in an appendix you pull up only if asked.

What each slide must actually do

Problem — make the audience feel the pain. Concrete, specific, quantified. Avoid "the world is inefficient." Whose day is ruined, and how much does it cost them?
Why now — the most-skipped, most-powerful slide. What changed that makes this the moment? No "why now," and you're a vitamin, not a painkiller.
Market — build bottom-up: (number of customers) × (price) = SOM, then widen to SAM and TAM. Top-down "1% of a $50B market" reads as lazy and gets discounted.
Traction — show the line going up and to the right: revenue, users, retention, pipeline, signed LOIs — whatever is most real. Pre-traction, show evidence of demand (waitlist, pilots, design partners).
Competition — never claim "no competitors" (means no market or no homework). Use a 2×2 or feature matrix; be honest about where rivals are strong; then show your wedge.
The ask — specific number, the milestones it funds, and the runway it buys. Tie raise to the next inflection (e.g. "to get to $1M ARR and a Series A").

Narrative & story principles

One idea per slide. If a slide makes two points, split it.
Headline = the takeaway, not the topic. "Revenue grew 4× in 12 months," not "Revenue." A reader should get the deck from headlines alone.
Set up → tension → resolution. Problem creates tension; solution and traction resolve it. Keep the arc; don't reorder into a feature catalog.
Concrete beats abstract. A named customer and a real number outsell adjectives.
Know your audience. Investor decks sell upside and defensibility; board/strategy decks sell focus, risk, and the plan. Same facts, different emphasis.
The deck supports the talk — it isn't the script. Slides carry the visual evidence; you carry the story.

Design do / don't

One idea, one chart, lots of whitespace.
A headline that states the point; the body proves it.
Consistent fonts, colors, and number formatting throughout.
Charts over tables; a labeled trend line over a wall of figures.
Big, legible type — assume it's read on a phone or projected across a room.

Don't

Paragraphs. If it's a paragraph, it's a doc, not a slide.
Five messages and three charts fighting on one slide.
Jargon, buzzwords, and unexplained acronyms.
Tiny fonts, dense tables, or charts with no axis labels.
Decorative clip art and effects that add nothing.

Financial model — proving the story

The model behind the deck need not be elaborate, but it must be honest and internally consistent. Investors probe the assumptions, not the spreadsheet's polish.

Unit economics (the heart of it)

Get these right before anything else — they decide whether growth creates value or burns it.

CAC (Customer Acquisition Cost) = total sales & marketing spend ÷ new customers acquired, over the same period. Include all the spend, not just ad cost.
Gross margin = (revenue − cost of delivering it) ÷ revenue. The share of each dollar you actually keep. Software is high (70–90%); services/hardware much lower.
LTV (Lifetime Value) = (average revenue per customer per period × gross margin) ÷ churn rate. Using gross margin, not raw revenue, is what makes LTV honest.
LTV:CAC — a rule of thumb is ≥ 3:1. Below ~1:1 you lose money on every customer; far above 3:1 may mean you're underinvesting in growth.
CAC payback = CAC ÷ (monthly gross-margin revenue per customer). How many months to earn the acquisition cost back. Under ~12 months is healthy for most businesses.

A minimal revenue build

Build revenue bottom-up from drivers, not a top-down "we'll hit $10M." Drivers you can defend; a target you can only assert.

new customers/month  ×  price (ARPU)            = new revenue/month
  + retained customers (prior − churn)          = recurring base
  → total revenue/month → quarter → year

Make the assumptions explicit and editable: acquisition rate, price, churn, expansion. The whole model should flex when you change one input cell — that's how you stress-test it.

Burn & runway

Burn = monthly cash out − cash in (net burn). Track gross burn too if revenue is lumpy.
Runway = cash in bank ÷ monthly net burn = months until you're out of money.
Tie the raise to runway: a round should buy ~18–24 months to the next milestone, with a buffer. State explicitly what the money achieves before you'd need to raise again.

Sanity-checking assumptions

Benchmark against reality — do your conversion, churn, and growth rates resemble comparable companies, or do they imply you'll be the best ever? Justify outliers.
Bottom-up vs top-down — if your bottom-up build and your market slide imply wildly different scale, one is wrong. Reconcile them.
One growth lever at a time — a model that needs CAC down and price up and churn down and market share up simultaneously is a hope, not a plan.
Run a downside case — what if growth is half and CAC is double? If the company dies instantly, the plan is too fragile. Show you've thought about it.
Cross-foot the numbers — totals add up, percentages have a base, units are consistent, the cash on the model matches the runway you claim.

Minimal model structure

ASSUMPTIONS (one tab/block, every editable input lives here)
  starting customers, monthly new, price/ARPU, monthly churn %,
  gross margin %, CAC, fixed costs (payroll, infra, overhead), starting cash

REVENUE BUILD (driven entirely by Assumptions)
  customers (begin + new − churned) → revenue → gross profit

COSTS
  S&M (implies CAC × new customers) + R&D + G&A + fixed

CASH / RUNWAY
  beginning cash + gross profit − costs = ending cash; net burn; months of runway

DERIVED METRICS (auto-calculated, never typed in)
  CAC, LTV, LTV:CAC, payback months, gross margin, MoM growth

Slide-by-slide checklist

One-line description a stranger understands in 5 seconds
Problem is specific, painful, and quantified
Solution is shown (screenshot/demo), not just described
"Why now" names a concrete, recent shift
Market sized bottom-up; SOM realistic
Business model + unit economics on a slide (LTV:CAC, payback)
Traction slide leads with the strongest real number
GTM is repeatable and affordable, not "it'll go viral"
Competition acknowledged honestly; your edge is clear
Team slide answers "why you?"
Ask states amount, milestones funded, and runway bought
Every headline states a takeaway, readable on its own
Model assumptions are explicit, editable, and survive a downside case
Deck and model agree on the numbers