Find a Blue Pond, Not a Blue Ocean

What Everyone Says

Go after a big market. Every pitch deck has the slide: a giant TAM, a land grab, a plan to own the category. VCs push for it because fund math needs outliers. Founders repeat it because a big market feels like safety. If the pond is huge, surely there's room for you.

Alon Talmor believed it too. His company Ask-AI was building an AI assistant for the whole enterprise, a market as big as software itself.

Why That's Wrong

A big market attracts the biggest guns. When ChatGPT launched, Ask-AI was close to $1 million in ARR, and Alon watched the foundation model labs raise billions to build exactly what he was selling. His words: "If we go for the enterprise AI assistant, they're gonna eat our lunch."

The hidden assumption in "go big" is that you'll be alone long enough to grow. In AI, you won't. The bigger and more obvious the market, the faster the giants arrive, and they arrive with more money than you will ever raise.

What Alon Did Instead

He went looking for what he calls a blue pond: "find a blue pond and make it blue." Not a blue ocean. A pond. A market small and hard enough that the giants don't want it, but valuable enough to build a real company in.

For Mosaic (the rebranded Ask-AI), that pond was technical support: the support teams at B2B software companies who replicate bugs, read log files, and need to know the product deeply. Conversational AI vendors chasing B2C-style support stay end-user facing and never go that deep. The problem was underserved precisely because it was hard.

His test for a market that looks crowded: maturity, not category. "Yeah, Salesforce can eat Gong's lunch, but yet Gong exists." When a market matures, sub-segments split off and specialists win them. Transactional support and technical support sound the same; they're different businesses.

The result: Mosaic serves about 100 enterprise customers on six-figure contracts, and the company is at multiple seven figures in ARR, approaching eight.

The Principle Underneath

Defensibility for a small company comes from depth the giants won't match, not breadth they can't. A blue pond has three properties Alon named: it's underserved, buyers genuinely need it and will pay for provable ROI, and it's a stepping stone toward your bigger vision rather than a dead end. Technical support checked all three. It also forces your product to learn everything about the customer's company, which compounds.

Should You Do This?

Shrink your market if you're competing on a roadmap that a far better funded company will inevitably ship, or if your category's giants stay shallow where your buyers need depth. Skip it if your "pond" is just a smaller version of the same shallow problem; the giants' feature release will still cover it.

The question to ask: in my market, what does everyone avoid because it's too hard? That's where the pond is. Hard is the moat.