Problem, Solution & Why Now with AI
The first three "real" slides of a pitch deck — Problem, Solution, and Why Now — decide whether the rest of the deck gets read. VCs frequently say that decks die in the first 90 seconds, and that 90 seconds is spent on exactly these three slides.
In this lesson you will learn the prompt patterns that produce sharp, specific, investor-grade Problem / Solution / Why Now content — even on the first try.
What You'll Learn
- The "specific person, specific pain" rule for problem slides
- How to write a one-sentence solution that does not sound like marketing fluff
- AI prompts to generate three Why Now candidates and rank them
- A red-team prompt that catches generic problem statements
Slide 1: The Problem
A weak problem slide says: "Companies waste billions on inefficient processes."
A strong problem slide says: "A 5,000-person regional bank spends 14 weeks on every new vendor onboarding. Of that, 9 weeks is paperwork passed between five teams. The Head of Procurement spends 40 percent of her time on status calls about that paperwork."
The difference: a real person, a real workflow, real numbers, and a real pain.
The Problem Slide Prompt
Audience brief (from Lesson 3): [paste]
I am writing the Problem slide of a pitch deck. My current rough version: [2-3 sentences].
Rewrite this in the "specific person, specific pain" format. Replace generic claims ("companies waste billions") with one specific buyer persona doing one specific workflow, the time or money it costs them, and the emotion that creates. Output 3 candidate versions ranked best to worst, and explain why the top version is sharpest.
The "rank 3 candidates" pattern is the secret. It forces the model to generate options rather than commit to its first guess.
Red-Team the Problem Slide
After AI gives you the version you like, run this:
Critique the Problem slide above as if you were a top-tier VC who reads 50 decks per week. Specifically: (1) does it pass the "would I bet $1M of my own money this pain is real" test, (2) is the persona specific enough that the audience can picture them, (3) is there a real number or is this hand-waving? Then rewrite the slide so it passes all three tests.
This second pass catches the soft language AI tends to slip in. Run it every time.
Slide 2: The Solution
The solution slide is where founders bury their pitch in product-feature soup. The fix: a one-sentence claim, then three supporting bullets, and a screenshot.
A weak solution slide says: "Our AI-powered platform leverages cutting-edge technology to revolutionize the way enterprises manage their workflows."
A strong solution slide says: "We replace the 9-week paperwork phase of vendor onboarding with a 3-day self-serve flow that uses AI to fill compliance forms automatically."
The strong version explicitly names what shrinks (9 weeks to 3 days) and what mechanism makes it shrink (AI-filled compliance forms). No buzzwords. No "leverage."
The Solution Slide Prompt
Given the Problem slide above, write the Solution slide in this exact structure:
(1) ONE sentence that states what we replace and what we replace it with. Format: "We replace [old thing] with [new thing]." Maximum 18 words. (2) THREE supporting bullets — each names a feature and the outcome it produces. Format: "[feature name]: [outcome with a number]." (3) A 1-sentence description of the ideal screenshot or product visual for this slide.
Avoid: "leverage," "revolutionary," "powered by," "next-generation," "platform that," "solution that."
The "avoid" list at the end of the prompt is critical. AI is very good at generating these filler words by default. Explicitly banning them forces concrete language.
Banned Word List
Add this banned list to every Problem and Solution slide prompt:
Banned words and phrases:
- leverage / leveraging / leveraged
- revolutionary / revolutionize
- cutting-edge
- next-generation / next-gen
- powered by AI (use "uses AI to [verb]" instead)
- platform that / solution that
- seamless / seamlessly
- empower / empowering
- one-stop shop
This list typically cuts the AI's filler tendency by 80 percent. It is one of the highest-ROI lines you can include in any deck prompt.
Slide 3: Why Now
Why Now is the slide most founders forget. VCs explicitly look for it. The question they are asking: "If this idea is so good, why has nobody built it before, and why is now the moment?"
A good Why Now answer points at a recent, verifiable, irreversible shift in the world. Examples:
- "Until late 2024, the regulator did not allow X. They lifted the restriction in November 2024."
- "GPU cost per million tokens fell 92 percent between 2023 and 2026, making our unit economics newly viable."
- "Remote inspection workflows became permanent post-2022, opening a market that did not exist before."
A weak Why Now points at vague trends: "AI is exploding," "post-COVID world," "digital transformation."
The Why Now Generation Prompt
Given the Problem and Solution above, generate 5 candidate "Why Now" answers. Each must point at a specific, datable, verifiable shift in the world from the last 36 months — not a general trend. For each candidate give: (1) the one-sentence claim, (2) the specific event or data point that proves the shift, (3) the year the shift happened, (4) how it makes our solution newly viable.
Rank the 5 candidates by how compelling each would be to a sophisticated VC. Then expand the top-ranked candidate into a complete Why Now slide.
The "5 candidates, rank them" pattern is gold. Most founders settle for their first Why Now. AI will generate 5 in 20 seconds; pick the strongest.
Putting Slides 1-3 Together
After you have all three slides, run one final coherence prompt:
Below are my Problem, Solution, and Why Now slides. Audit them together. Specifically:
- Does the Solution actually solve the Problem as stated? Or does it solve a different problem?
- Does the Why Now actually make the Solution newly viable? Or could the Solution have existed 5 years ago?
- Is the same persona present in all three slides?
- Are there any contradictions between slides?
Flag anything weak and propose a revised version of the weakest slide.
This audit prompt catches the single most common mistake in pitch decks: a Problem about Persona A, a Solution that serves Persona B, and a Why Now about an unrelated trend.
A Worked Mini-Example
To make this concrete, here is what the full Lesson 4 workflow produces for a hypothetical startup:
Problem (after the rank-3 prompt + red-team pass)
Heads of Procurement at $500M-$2B companies spend 14 weeks onboarding each new vendor. Nine of those weeks are paperwork passed between procurement, legal, security, and finance. Their teams ship 30 percent fewer vendor relationships per year than the business asks for.
Solution (after banned-words prompt)
We replace the 9-week paperwork phase of vendor onboarding with a 3-day self-serve flow.
- Compliance form AI: auto-fills 80 percent of the forms vendors must submit
- Approval routing: cuts internal hand-offs from 5 to 1
- Audit log: drops audit prep time from 6 weeks to 4 days
Why Now (after 5-candidate rank prompt)
The Federal Trade Commission's 2024 Third-Party Risk Rule made automated vendor risk monitoring mandatory for regulated industries. Procurement teams now have a hard compliance deadline of Q3 2026 to deploy tooling, and most have no incumbent solution.
Notice how the three slides reinforce each other: same persona, same pain, same time horizon, no buzzwords.
Key Takeaways
- Problem slides must name a specific persona, specific workflow, and specific number — not "companies waste billions"
- Solution slides should follow: 1 sentence of replacement, 3 outcome-named bullets, 1 visual
- Always include a banned-words list in your prompt to kill "leverage" and "revolutionary"
- Generate 5 Why Now candidates and rank them — the first one is almost never the strongest
- Run a coherence audit on Problem / Solution / Why Now together before you build any more slides

