How to Validate a SaaS Idea: The 4-Step Discovery Process We Use With Founders
Most founders spend between $50,000 and $150,000 building SaaS products nobody wants. We often tell founders who come to us not to build yet. Not because we don’t believe in them. Our job isn’t to build your idea; it’s to tell you the truth about it.
At SMELighthouse, we’ve validated over 200 SaaS ideas. The founders who succeed all pass the same four gates, while those who skip them usually end up building the wrong product.
This article breaks down the discovery process we use with early-stage founders —the smoke tests, the hard questions, and the real signals that tell us whether an idea deserves time, capital, and commitment.
The SMELighthouse Discovery Process
Our validation sprint takes between two to four weeks. Ideas die here sometimes, and that’s the point. We break it into four steps executed sequentially. Skip one and you’ll build the wrong thing faster.
Step 1: Problem-Solution Fit
We start with the Smoke Test Principle. Before you write a line of code, gauge interest from the people you’re building for. In B2B SaaS, this means getting one or two target companies to signal interest through a letter of intent or preliminary contract that says if you build this, we’ll use it and pay for it.
The holy grail is they pay you before you build. That way you’re using client money to execute. But that’s rare. The next best thing is a signed commitment.
Before engaging in that conversation, it’s important to understand who you’re speaking with. Most founders conflate buyer persona (who signs the check) with user persona (who uses the product). In B2B, they’re often different.
Take a compliance SaaS we validated for example. The buyer was the Chief Risk Officer, who cared about regulatory fines and had a budget, while users were IT (server access), HR (employee PII), and legal (contract review). The founder initially targeted IT but the CRO was the decision maker. And that insight changed the entire product pitch and feature set.
The 5-Question Problem Drill
We run this drill in the first call. If a founder can’t answer these without mentioning their solution, we ask them to think more deeply and don’t proceed to the next step until they have a sense of what the answers to these are:
- What job is the user trying to get done? (Functional + emotional)
- How do they do it today? (Workarounds = gold)
- What happens if this problem goes unsolved? (Urgency test)
- Who else cares about this problem? (Stakeholder map)
- Would they pay to make it vanish? (Willingness to pay)
Red flag: If the answer to number three is “nothing really,” you don’t have a business idea. Either pivot or kill it.
Step 2: Market Demand Signal
Once you understand the problem, you need proof that people want the solution. One common approach is to run surveys of your target audience. While surveys can give you signals, they don’t always predict real behavior. Someone saying “yes” on a survey doesn’t necessarily mean they’ll act when a purchase decision is required. In many cases, surveys hide true intent because there’s no cost to agreement.
We run a fake door test before a single wireframe, using a Landing Page Smoke Test. Build a one-pager with problem statement, value prop, and “Join Waitlist.” Drive $500 in targeted ads to your ICP. Then measure conversion rate:
< 5% conversion = No demand. Kill it.
5–10% = Explore further
10% = Greenlight
Tool Stack: Unbounce + Google Ads + Calendly
Cost: $500–$1,000
Timeline: 7 days
A fintech founder targeting mid-market CFOs got 12% conversion. One CFO wrote in the Calendly notes, “If this works, I have 5 other CFOs who need it.” We greenlit immediately and he had his first 3 LOIs before writing code.
The strongest signals of demand come from getting users to go through a real decision flow such as a $0 credit card check or waitlist signup because it shows actual intent, not just agreement.
Step 3: Solution Clarity & Scope
Our second principle is start small and scale from there. Founders who’ve been in an industry for 10–15 years often want to build everything they know. But we fight that instinct. To do this, we validate cheaply and fast by using AI to create mockups and generate high-fidelity prototypes in hours for rapid testing.
It’s important to note that AI doesn’t replace the need for user testing; it accelerates it. You can generate multiple versions of a feature set, test which resonates, and iterate based on real feedback. The key is using AI to fail faster, not to avoid failure altogether. AI gives you speed. User testing gives you the truth.
We run five usability tests with target users on these AI-generated mockups and measure if they can complete the core task and where they get stuck. If the success rate is below 60 percent, the scope is wrong.
A founder came to us with a complex fintech idea that would have taken months to prototype traditionally. Using AI-generated mockups, we had testable screens in two days and put them in front of potential users immediately. The feedback revealed that the core workflow was flawed. We pivoted the design before spending a single dollar on development.
Step 4: Financial Validation
Most founders start financial validation with TAM (Total Addressable Market—everyone who could buy your product), SAM (Serviceable Addressable Market—the portion you can realistically reach), and SOM (Serviceable Obtainable Market—the portion you can actually capture).
We go a level deeper by running price anchoring and willingness-to-pay interviews.
We use the Van Westendorp Pricing Test.
At what price is this product too cheap, suspicious, good value, expensive but worth it, or too expensive?
Kill criteria: If zero percent of interviewees say it’s expensive but worth it, you have a vitamin, not a painkiller.
A founder thought $50 per month was fair. Interviews revealed compliance officers saw $50 as cheap, which probably doesn’t work. At $200 per month, they said expensive but worth it if it saves 10 hours per week. We priced at $180 and hit $10K MRR in three months.
Before You Build: Common Validation Mistakes
Can you validate a SaaS idea without writing code?
Yes.
90% of validation happens before code. Our most successful client used AI mockups and a waitlist page to get three LOIs before hiring a developer.
How long should SaaS validation take?
Two to four weeks if you’re disciplined.
Six to eight weeks if you’re learning the process.
Longer than that and you’re stalling.
How do I know if my SaaS idea is worth building?
You can answer these four questions with data.
- Do at least 5 target customers confirm this is a top 3 problem?
- Did your smoke test get above 10% conversion?
- Did your prototype hit 60% plus usability success?
- Did 20%+ of interviewees say expensive but worth it?
If yes, build. But if no, pivot or kill.
When We Tell Founders Not to Build
There are clear patterns that signal when an idea isn’t ready for development. These aren’t arbitrary judgments; they’re data-driven decisions based on signals we’ve seen across hundreds of validations.
Scenario 1: The Problem Isn’t Painful Enough
When responses to the 5-Question Problem Drill reveal that the problem ranks low on urgency, that’s a red flag. If users say the issue is nice to have or we’ll address it next quarter, you’re building a vitamin, not a painkiller.
We see this often in efficiency tools. Founders pitch automation that saves 30 minutes per week and users shrug. The ROI isn’t compelling enough to justify budget or behavior change.
Scenario 2: Buyer and User Persona Misalignment
Founders frequently target the wrong stakeholder. They build for the user who will love the product but ignore the buyer who controls the budget.
When our stakeholder mapping reveals that the economic buyer doesn’t feel the pain directly, and the user has no purchasing power, we recommend a pivot. We’ve seen founders spend months building a tool that IT loves but procurement won’t approve because it doesn’t address a business-level problem.
Scenario 3: The Smoke Test Fails to Generate Interest
A landing page with a waitlist that converts below 5% after targeted ad spend is telling you that the market isn’t interested.
We’ve had founders argue that people just don’t understand yet. They want to spend more on education, while we recommend pivoting or killing it. The reason is that education is expensive, but market readiness is cheap.
Scenario 4: Feature Bloat Without User Clarity
When founders want to build everything they’ve imagined over 10–15 years in an industry, we push back.
If prototype testing shows users getting lost in a 12-feature dashboard, we don’t recommend adding tutorials. We recommend removing nine features. Founders resist this because they believe users need the complexity, while data shows otherwise.
Scenario 5: The Pricing Test Reveals No Premium Value
When 100% of interviewees say your proposed price is about right or cheap, that’s not good news. It means they don’t see enough value to justify a premium.
The Van Westendorp test should reveal a segment willing to pay “expensive but worth it.” If that segment is zero, you haven’t found a painful enough problem.
The Common Thread
In every scenario where we tell founders not to build, the underlying issue is often the same. They’ve fallen in love with their solution before falling in love with the problem.
Validation exists to break that infatuation. And sometimes that means killing the idea, while sometimes it means killing the founder’s ego. Both are expensive. But both are cheaper than building the wrong product.
FAQ: SaaS Validation Questions From Founders
Q: What is the first step in SaaS validation?
If you can’t articulate the problem without mentioning your solution, you don’t have a business.
Q: How much does a SaaS validation sprint cost?
A: Cost depends on scope and timeline.
Most founders budget between two to four weeks of focused work..
Q: Can SMELighthouse validate my SaaS idea if I’m pre-revenue?
A: Absolutely.
Most of our founders are pre-revenue. That’s when validation matters most.
Q: What happens if my SaaS idea fails validation?
A: You pivot or kill it.
We help you find the pivot. One founder pivoted three times in our sprint and landed on a problem worth $2M ARR.
Q: How is AI used in SaaS validation?
A: We use AI to generate mock-ups, analyze interview transcripts for patterns and simulate user flows.
AI doesn’t replace customer conversations. It accelerates them.
Final Take: Why Validation Is Your SaaS Compass
Validation is not a gate you pass through to get permission to build. It’s a compass that tells you whether to build at all.
Most founders treat validation as a box to check; we treat it as the business. The best SaaS ideas don’t need to be built. They need to be proven.
If you’re sitting on a SaaS idea, book a 30-minute validation review with our team. We’ll pressure test it together.
Additional resource
Maurya, A. (2012).Running Lean. Smoke test methodology
