How to Validate a SaaS Idea Before You Write Code

The most common way a micro-SaaS fails is not bad code. It is building something for months and discovering that nobody will pay for it. AI coding tools make this mistake easier, because building feels cheap. Validation is the work you do before building to find out, with evidence, whether a specific group of people has a painful problem and will pay you to solve it. Here is a practical, four-step process that takes two to four weeks.

Step 0: Score your ideas first

Before talking to anyone, score each idea from 1 to 5 on these factors:

Ideas scoring 30 or more out of 40 are worth validating. Below that, rework the idea or the audience.

Step 1: Problem interviews

Talk to 10 to 20 people who have the problem. Recruit them from your network, industry communities, LinkedIn, or businesses you can visit. Offer a small thank-you, such as a gift card or a free year of the product.

Ask about the past, not the future

People are poor at predicting what they would buy. They are good at describing what they actually did.

```

1. Walk me through the last time you had to [task].
2. How often does that happen?
3. How do you do it today? What tools or people are involved?
4. What does that cost you, in money or time?
5. What is the most frustrating part?
6. Have you tried other tools? Why did you stop using them?
7. If you could wave a magic wand, what would happen instead?

```

Avoid: "Would you use a tool that...?" and "How much would you pay for...?" Those questions invite polite yeses.

Signals to look for

Keep notes in a simple spreadsheet: who, role, how often, current solution, cost, pain score from 1 to 5, and quotes in their own words.

Step 2: A landing page with a real price

Build a one-page site that explains:

• Who it is for (be specific).
• The outcome in plain words, using phrases from your interviews.
• How it works in three steps, with a mock-up or screenshots.
• A real price.
• A clear call to action: join the early access list, book a demo, or pre-order.

Drive targeted traffic: personal outreach to interviewees and their colleagues, posts in communities where self-promotion is allowed, and perhaps a small ad test aimed at your exact audience.

Benchmarks (rough, vary by channel)

Cold, untargeted traffic converts far lower. Judge the page on traffic from the people you actually want to serve.

Step 3: Pre-sell

Money is the clearest evidence. Ways to ask for it:

• Founding member pricing: for example, the first year at 50 percent off, paid upfront.
• Paid pilot: a fixed fee for 60 to 90 days of a hands-on version.
• Deposit: a refundable deposit to reserve a place at launch.
• Letter of intent: for larger B2B deals, a signed commitment to buy if the product meets stated requirements.

Use a simple payment link to collect payment, state your delivery date, and promise a full refund if you do not deliver. Honour it.

Tip: If people say "I love it" but will not pre-order, ask what would need to be true for them to pay today. The answer tells you what is missing: a feature, trust, a budget cycle, or the wrong buyer.

Step 4: A concierge or no-code MVP

Before writing much code, deliver the outcome by hand or with no-code tools. If your idea is an automated weekly report, build the first few reports manually in a spreadsheet and email them. If it is a scheduling tool, run the scheduling yourself with existing apps.

This teaches you:

• What customers actually use and ignore.
• What "done" looks like for them.
• Where the edge cases are.
• Whether they keep paying after the novelty wears off.

Worked example

A founder who spent six years as a dispatcher believes small HVAC companies waste hours turning technician notes into customer invoices.

• Interviews: 14 HVAC owners and office managers. Eleven say invoicing takes 4 to 8 hours a week; seven pay a part-time bookkeeper partly for this; three ask when they can try a solution.
• Landing page: "Turn tech notes into invoices in 2 minutes", priced at $79 a month. 180 targeted visitors from outreach and an HVAC forum produce 41 signups (about 23 percent) and 12 calls.
• Pre-sales: 6 companies pay $399 for a 90-day paid pilot, totalling $2,394.
• Concierge MVP: for the first month, the founder converts notes into invoices with a template and an AI assistant, reviewing every one, before building the software.

Six paying pilots, clear pain and a known workflow: that is enough evidence to start building.

When to change course

If 20 targeted conversations produce no pre-orders, no paid pilots and no urgent pain, change one variable at a time:

1. The audience: a different industry or company size with the same problem.
2. The problem: a more painful adjacent problem you heard about in interviews.
3. The offer: a service or done-for-you version before software.
4. The price or packaging: sometimes the problem is real but the offer does not match how they buy.

Warning: Friends, family and other founders are the least reliable validators. They want to be supportive. Talk to strangers who have the problem and the budget.

Validation checklist

• ☐ Idea scored 30+ out of 40
• ☐ 10 to 20 problem interviews completed and logged
• ☐ Clear evidence of current spending or time cost
• ☐ Landing page with a real price live
• ☐ Targeted traffic sent and conversion measured
• ☐ At least 5 to 10 pre-orders, deposits or paid pilots
• ☐ Outcome delivered manually to early customers
• ☐ Decision made: build, adjust or move on

Your next step

List ten people who have the problem you want to solve and book three conversations this week. Use the interview questions above and write down exactly what they say.

The full guide, Building and Selling a Micro-SaaS, continues from validation through building with AI coding tools, SaaS architecture, Stripe subscription billing, pricing, SaaS sales tax and privacy, getting the first 100 customers, metrics and selling the business, for $0.99.