How to Validate a Business Idea in 90 Days

From "Aha!" to Action: Your First 90 Days

You’ve had the jolt—the business idea that won’t let you sleep. Good. Now forget the romance of the "ideation phase" and get to work. Most ideas don't fail because they were bad. They fail because the founders didn't systematically de-risk them.

This isn’t about writing a 50-page business plan. It’s a 90-day sprint to validate your core assumptions and prove your idea deserves to exist. We'll break it down into three phases: De-risking, customer discovery, and building proof.

Phase 1: De-Risking the Idea (Weeks 1-4)

Your goal this month is to kill the idea if it’s not viable. That means understanding the competitive landscape, assessing the market size, and being honest about your right to win.

Is Anyone Else Doing This? Good.

Stop searching for a "unique" idea. Very few are. Your first step isn't to find an untouched market; it's to understand the one you're entering. Exhaustive research is non-negotiable.

Where to look: Go beyond a simple Google search. Dig through Product Hunt, Crunchbase, G2, Capterra, and industry-specific forums on Reddit or Hacker News. Find every existing or failed company that has attempted to solve this problem. · Common mistake: Dismissing competitors as "not good enough" or "missing the point." Your conviction is not a substitute for evidence. If their solution sucks and they still have customers, that's a powerful signal the underlying problem is real and painful. · The non-obvious insight: Competition is validation. The existence of competitors proves a market and a customer's willingness to pay. A complete absence of competition is often a red flag. It may mean the market doesn't exist, is impossible to monetize, or you're simply not searching correctly.

Define Your 10x Differentiation

It’s not enough to be slightly better. You need a defensible wedge that makes switching from an existing solution a no-brainer. Can you be:

10x cheaper? (Example: "We cut enterprise SaaS costs by 80% by leveraging open-source.") · 10x faster or more efficient? (Example: "Our design tool renders complex files in 5 seconds, not 50.") · 10x easier to use? (Example: "You can set up a full e-commerce backend in three clicks.") · Targeting a fundamentally underserved niche? (Example: "All existing CRMs are for sales teams; ours is built exclusively for freelance architects.")

Without a clear, compelling answer, your idea will drown in the noise.

Sanity Check: Is This a Venture-Scale Problem?

Venture capitalists look for businesses that can realistically generate >$100M in annual revenue. Not every good business is a venture-scale business. Run a quick back-of-the-envelope market sizing:

Total Addressable Market (TAM): Total number of potential customers x your potential annual price. (e.g., 1M businesses in the US x $10,000/year = $10B TAM). Be realistic. Is the market huge? If not, it may be a great lifestyle business, but it's not a fit for VC funding.

Phase 2: Customer Discovery & Validation (Weeks 5-8)

Stop refining your pitch deck and start talking to humans. Your goal is to conduct 20-50 "problem interviews" with your target customer profile. Do not try to sell them anything.

Finding Your First 20 Conversations

Identify your ideal customer profile (e.g., "VPs of Engineering at 50-200 person tech companies"). Now, go find them.

LinkedIn: Use Sales Navigator for precise filtering. Send a direct, honest, and non-salesy message. · Online Communities: Find where your customers hang out. Is it a specific subreddit, a professional Slack group, or an industry forum? Participate genuinely before you ask for help. · Your Network: Ask friends, former colleagues, and advisors for warm introductions.

"Hi [Name], I'm exploring solutions to [problem space] for [customer profile, e.g., engineering leaders]. I'm not selling anything. I'm just trying to learn from experts like you about how you currently handle [the problem]. Would you be open to a 15-minute chat to share your experience?"

The Problem Interview Script

Your only goal is to learn. Ask open-ended questions about their current workflow and pains. Shut up and listen.

Context Questions: "Can you walk me through how you currently [do the thing your idea relates to]?" · Problem Questions: "What's the hardest part about that process?" "What tools or workarounds are you using?" "Have you ever tried to search for a better solution?" · Consequence Questions: "What happens if you fail to solve this? How much time or money does that cost you?"

"That sounds like a neat idea." (Translation: This is not a real problem for me.) · "Yes, I would definitely use that!" (People are polite. This is not a commitment.) · They aren't already using some kind of workaround. (If the problem isn't painful enough for them to have cobbled together a solution—even a bad one—it's not a painkiller.)

You're looking for a "hair on fire" problem. If you hear things like, "We hacked together three spreadsheets to do this, and it takes 10 hours a week and my team hates it," you're on the right track.

Phase 3: Building Proof & Getting Commitment (Weeks 9-12)

You’ve validated the problem. Now you need to validate your proposed solution and get tangible proof of interest before you write a line of code or hire an engineer.

The Minimum Viable Test (Not Product)

Your goal is to build the smallest possible thing to test your core value proposition. It is not a feature-rich MVP. It could be:

A Landing Page: Describe the value proposition and have a single call-to-action ("Request Early Access"). Drive traffic to it with a few hundred dollars in ads. Does anyone sign up? · A Figma Prototype: A clickable, high-fidelity mockup of your app. "Sell" this prototype in your next round of customer calls. Can you get them to a "wow" moment? · A "Concierge" Service: Manually deliver the promised value. If your idea is an automated reporting tool, start by manually creating the reports for your first 5 users. This forces you to learn the exact steps and prove the value before you automate.

Ask for Commitment, Not Compliments

The only true validation is when a customer gives you something of value. This is how you separate the polite "maybes" from real early adopters. Ask for:

Time: "Would you be willing to commit to a 30-minute feedback session with me every two weeks as I build this?" · Reputation: "This is fantastic. Is there anyone else on your team, or your boss, who I should show this to?" · Money: This is the ultimate test. You can ask for a pre-payment for the first three months of service or, for larger enterprise customers, a non-binding Letter of Intent (LOI).

An LOI is a powerful signal. Getting 5-10 of these from reputable companies—stating they intend to buy your product for $X if you build Y—is often enough traction to raise a pre-seed round of $500k to $2M.

What Now? The Path Forward

After 90 days, you should have a clear verdict. You've either invalidated your assumptions and saved yourself years of wasted effort, or you have the early proof points you need to take the next step. That next step is a choice: bootstrap the business using early revenue, or raise a pre-seed round to accelerate your path to building a product and finding product-market fit. Whatever you choose, you're no longer operating on a blind guess. You're operating on evidence.

How to Apply This This Week

Create a competitive matrix. In a spreadsheet, list 5-10 competitors and map them by target customer, pricing, and their core differentiation. Be brutally honest about where you fit. · Draft your "problem-seeking" outreach message. Write a concise, non-salesy template you can use on LinkedIn or email. · Identify 10 people you want to interview. Use LinkedIn or your personal network to build your first target list. · Send 5 outreach messages. The journey of 50 interviews begins with a single send. Get started now.