How to Validate a SaaS Idea Before Spending $35,000
Not every idea deserves an MVP. Here's how to figure out if yours does — before writing a check.
The Three Questions
Before we take on any project at M8 Apps, we ask:
1. Is someone already paying to solve this problem?
If people are paying for spreadsheets, manual processes, consultants, or competitor products to solve this problem — there's a market. If nobody is spending money on this today, you'll need to create demand, which is 10x harder.
2. Can you reach 100 potential customers in 30 days?
Not hypothetical users. Real people you can email, call, or meet. If you can't get 100 people to look at your product in a month, distribution is going to be your bottleneck — not the product itself.
3. Would they pay before you build?
The strongest validation signal is a pre-sale or deposit. Create a landing page, describe the product, and ask for a commitment. If 5-10 people put down a deposit on a product that doesn't exist yet, you have something real.
The Discovery Sprint as Validation
Our $4,500 Discovery Sprint is designed as a validation tool. In 1-2 weeks, you get:
- A clear product scope (what to build and what NOT to build)
- User flows that show the actual experience
- Technical architecture that proves it's feasible
- A fixed estimate so you know the real cost
If Discovery reveals the idea needs fundamental changes, you've spent $4,500 instead of $35,000 learning that lesson.
Validation Signals We Look For
- Strong signal: Pre-sales, waitlist signups, letter of intent from a customer
- Good signal: Paying customers on a manual/scrappy version of the product
- Weak signal: "Everyone I talked to said they'd use it" (talking ≠ paying)
- No signal: "I think this is a great idea" (your opinion is not data)
When to Skip Validation and Just Build
If you have domain expertise, existing customers, and a clear pain point you've observed for years — skip the surveys and build. Nutricionista.ai was built because the nutritionist had been doing the work manually for a decade. She didn't need a survey. She needed software.
Have an idea worth validating? Let's talk about it.