Dex raised $5.3 million by solving one expensive problem: hiring AI engineers. The original post makes a clean argument — pick a narrow market with high pain, scarce supply, and clear signal. Build for that slice. Own it. The logic is tight.
But if you're a founder reading this, you probably had the same thought I did: That works when your niche has 15,000 engineers and 50 paying customers in a few months. My market doesn't look like that.
You're right. Not every founder can afford to niche down like Dex did. The objection is real. Let me take it seriously and walk through when narrow focus works, when it doesn't, and what to do if you're stuck in the middle.
The objection: narrow markets can be too small to sustain a business
Dex picked AI engineers. That's a market with thousands of companies hiring, each willing to pay 20-30% of a $200K+ salary. The math works: 50 customers at $36K average fee = $1.8M ARR. Fast.
But what if your niche is something like "recruiting for geothermal drilling engineers"? Or "sales tools for pet insurance brokers"? The total addressable market might be 200 companies globally. Even if you win every single one, you're capped at a lifestyle business. Venture capital won't touch it. And you can't raise $5.3M on that.
This is the real objection most founders face. The advice to "niche down" sounds great in a blog post. In practice, it feels like you're painting yourself into a corner.
When narrow focus actually backfires
Three scenarios where the Dex playbook breaks:
- Your niche has no repeat buyers. If your customers hire once every three years, you can't build a recurring revenue model. Dex wins because companies hire AI engineers every quarter. The buying cycle matters as much as the pain.
- Your niche is too price-sensitive. Dex charges 20-30% of salary because the cost of a bad hire is millions. If your market's bad outcome costs $10K, they won't pay $5K to avoid it. The willingness to pay has to match the pain.
- Your niche doesn't have clear signal. Dex evaluates candidates through coding challenges and published research. That works for AI engineers. It doesn't work for, say, creative directors or sales VPs, where evaluation is subjective and network-dependent.
If any of these apply to you, copying Dex's exact playbook will leave you with a handful of customers and no growth path.
The real lesson isn't niche selection — it's problem selection
Here's what I think the original post gets right, even if the niche advice doesn't fit everyone: Dex didn't start with a market. They started with a problem.
Paddy Lambros spent two and a half years watching startups stall because of bad AI hires. He didn't say "I want to build a recruiting company." He said "I want to solve the problem of bad AI hires." The niche was a consequence of the problem, not the starting point.
Most founders do the reverse. They pick a market ("I want to build a sales tool for SMBs") and then look for a problem inside it. That's why you end up building for "everyone who needs X" — because you haven't found the expensive problem yet.
So the question isn't "How narrow should my niche be?" It's "What expensive problem can I solve that has the three properties Dex found?"
How to find your expensive problem when your market feels small
If you're in a market that doesn't have 15,000 potential candidates and 50 ready buyers, here's a different approach:
- List every pain point your customers mention more than once. Not the feature requests. The complaints. The things they say "if only someone would fix this." Write them down. Rank by how much money a bad outcome costs them.
- Find the one where the cost of inaction is higher than the cost of your solution. This is the hair-on-fire test. If a customer can ignore the problem for six months without losing money, it's not expensive enough.
- Check if the problem repeats. Do they need to solve it every quarter? Every year? If it's a one-time fix, you need a different business model — consulting, not SaaS.
- Test willingness to pay before you build anything. Call five people in your target market. Say "I'm working on a solution to [problem]. Would you pay $X/month to have it solved?" If three say yes, you have something. If none say yes, your problem isn't expensive enough.
This is how we built MiraReach. We didn't start with "AI sales outreach platform." We started with the problem: founders spending four hours a day on prospect research and email drafting, and still sending generic messages that get ignored. The product came after the problem.
What to do if your niche is genuinely too small
Let's say you've done the work. You found an expensive problem. But the total addressable market is 500 companies, and you need 200 customers to hit $2M ARR. That's tight.
You have three options:
- Expand the problem, not the market. Instead of serving more customer types, serve the same customers with more solutions. If you're solving hiring for geothermal engineers, add onboarding or compliance. Increase revenue per customer instead of customer count.
- Find an adjacent market with the same problem. The pain of bad AI hires isn't unique to startups. It exists in hedge funds, defense contractors, and research labs. Dex could expand into those without changing their process. Your niche might have cousins you haven't looked at.
- Build a process, not a product. Dex's real asset isn't their software. It's their network of 15,000 engineers and their evaluation methodology. That process works for any technical hiring market. If your niche is small, build a repeatable process and license it to others who serve similar niches.
The third option is the one most founders ignore. They think they need to build a platform. Dex built a service that looks like a platform. The difference matters because services can pivot faster. If the AI engineer market dries up, Dex can apply their process to ML ops engineers or data scientists. The process is the moat, not the niche.
The takeaway: don't copy Dex's niche, copy their logic
The original post is right about one thing: most founders build for "everyone who needs X" and end up serving nobody well. But the solution isn't always to pick a narrower niche. Sometimes it's to pick a better problem.
Dex's success came from finding a problem with high pain, scarce supply, and clear signal. Those three properties are what matter. The niche is just where those properties happen to cluster.
If your market doesn't have 15,000 engineers, don't panic. Find the expensive problem. Test willingness to pay. Build a process that solves it. And if the market is genuinely too small, expand the problem before you expand the market.
That's how you ride the wave without copying the playbook.
If you're running your own pipeline and want to see how we apply this logic to sales outreach, give MiraReach a try. We built it for founders who need to find and close their first 50 customers — not for everyone who sends cold email.
— Mira