Marketing & SEO

Demand Validation

Run lightweight experiments to confirm real people want what you built — before investing heavily in marketing it.

Demand Validation

Run lightweight experiments to confirm real people want what you built — before investing heavily in marketing it.

Why It Matters

Marketing amplifies signal. If there's no demand signal, marketing just amplifies silence — faster and louder.

Demand validation establishes a baseline: do real people, with real money, have this problem urgently enough to take action? "My friends think it's a good idea" is not demand validation. Ten strangers signing up to a waitlist with their real email is.

Do this before investing in SEO, content, or distribution. A strong signal gives momentum. A weak signal tells you to pivot positioning or reconsider the product — and the sooner you know, the better.

The Hierarchy of Demand Evidence

From weakest to strongest:

  1. Someone said they'd use it — near worthless
  2. Someone signed up for a waitlist — weak signal (low friction)
  3. Someone gave their email after reading your pitch — moderate
  4. Someone engaged with your content about the problem — moderate
  5. Someone clicked "start free trial" — strong
  6. Someone entered a credit card — very strong
  7. Someone paid — definitive

Goal: Get to at least level 4-5 within a day before committing to a full marketing push.

Four Validation Experiments

1. Landing Page Smoke Test (2-3 hours)

A minimal landing page with a real CTA (email capture or "request access") measures response rate before full marketing investment.

What to build:

  • Headline (your positioning statement)
  • 3-sentence problem description
  • 3-bullet solution (outcome-focused)
  • Email capture or "Get early access" CTA

Success metric: 5%+ conversion from unique visitors to signups. Under 1% suggests a copy or positioning problem.

Build options: v0.dev + Vercel, Carrd, or Typedream.

2. Community Polls (1-2 hours)

Post in 3-5 subreddits, Discord servers, or Slack groups where your ICP hangs out. Ask about the problem — not your product.

"Quick question for [role] — how do you currently handle [specific problem]? Do you use [option A], [option B], or something else?"

Don't mention your product. You're listening. Note: how many respond (prevalence), language they use (copy gold), solutions they've tried (competitor intel), how urgent the problem sounds.

3. Direct Outreach (1-2 hours)

Identify 20 people matching your ICP via LinkedIn, Twitter/X, or Reddit. Send a genuine research question, not a pitch.

Target response rate: 20-30% with a research frame vs. a sales frame.

Even 3-5 responses give you qualitative patterns worth the effort.

4. Content Engagement Test (ongoing)

Post 3-5 pieces of content about the problem (not the product). Measure engagement.

Strong signals: Comments from people saying "I have this exact problem," DMs asking about your solution, shares from your target audience.

Use FastWrite to generate and optimize these posts without promoting your product directly.

Interpreting Results

Signal Interpretation
Strong engagement + DMs + 5%+ smoke test conversion Accelerate marketing investment
Moderate engagement + some signups Demand exists; refine positioning
Weak engagement + low conversions Positioning problem, wrong channel, or low-urgency problem
No response Problem may not be urgent enough, or you're in the wrong communities

Setting Validation Criteria Upfront

Define success before running experiments. Use MetricGen to model validation thresholds:

  • Minimum email signups to proceed with full marketing
  • Minimum outreach response rate to validate demand
  • Minimum content engagement to validate the channel

Having numbers upfront prevents rationalizing weak results as "good enough."

Deliverable

One markdown file: validation-results.md containing:

  • Which experiments you ran
  • Quantitative results
  • Key qualitative insights
  • Go / no-go decision
  • If go: top 3 validated pain points for copy
  • If no-go: hypothesis for what to change

Resources

  • MetricGen — Model validation thresholds and success metrics
  • FastWrite — Generate problem-focused content for testing
  • Carrd — Fast smoke test landing pages

Ready to build?

Go from idea to launched product in a week with AI-assisted development.