
The Total Addressable Market (TAM) and its siblings, the Serviceable Available Market (SAM) and the Serviceable Obtainable Market (SOM), are probably the most misunderstood, miscalculated, and misused data points for startup investors.
For many years, we've used TAM as a guide to the size of the market, searching for startups with large markets where they could compete, grow to scale, and eventually be acquired by larger companies. In many markets, this is still an important metric. But I'm finding that in the fast-paced world of AI companies, it's not just the size of the market that matters, but the company's ability to expand or create the market.
If you've ever read the classic business strategy book Blue Ocean Strategy, you know that highly competitive markets are "red oceans," where taking market share from incumbents is difficult and expensive. Blue oceans, on the contrary, are wide-open markets with little or no competition. The book describes ways that startups can innovate to carve out blue ocean opportunities, even in established red ocean markets.

Today's venture capital investors are looking at AI deals every day, trying to find promising companies addressing large TAMs. It can be a trap, however, to focus too much on TAM because many of the first-generation AI companies I see are simply applying AI to do what people have always done, only faster and cheaper.
That's great, but the moat is often shallower than it first appears. Fast followers will soon flood the market with me-too products, prices will fall, and the startup's competitive advantage may disappear.
What angel and VC investors should be looking for are companies that can expand the market by using AI to do something that wasn't previously possible, or wasn't economically feasible.

Consider ElevenLabs, the AI voice technology company. Traditional dubbing and translation of video content have always been expensive, so producers generally limited localization to their largest markets. Smaller markets often weren't worth the investment.
ElevenLabs is changing that equation. In 2026, Vimeo integrated ElevenLabs' AI dubbing technology, allowing its customers to translate videos into dozens of languages. Since January, Vimeo customers have dubbed more than 1.4 million minutes of video across more than 25,000 videos.

What's interesting isn't just that dubbing is now faster and cheaper. It's that thousands of videos can now reach audiences that would never have justified the cost of traditional localization. ElevenLabs isn't simply competing for a share of the existing dubbing market. It's helping create demand that previously didn't exist.
This is the kind of market expansion that should get investors excited.
Of course, simply creating a market isn't enough. The company still needs a defensible competitive advantage, a profitable business model, and the ability to capture a meaningful share of the value it creates. Otherwise, it may end up building a wonderful new market for its competitors.
As AI expands into virtually every area of technology, I believe the most valuable startups won't necessarily be those attacking the largest existing markets. They'll be the companies that use AI to create markets we haven't even thought of yet.
So perhaps the most important question to ask a startup isn't "How big is your TAM?" but "How much bigger can you make it?"


.jpg)


