There’s a number you’ll find in almost every investor presentation for a growth company.
TAM. It stands for “total addressable market.” In plain English, it’s the amount of money a company could theoretically make if it captured every dollar spent in its market.
The bigger the TAM, the longer the growth runway, and the greater the stock’s upside potential.
That makes TAM one of the first things investors look at when sizing up a growth stock.
And after studying some of the greatest growth stocks of the past two decades, I think investors routinely underestimate it.
They look at the market in front of them today and assume tomorrow will look roughly the same.
Today, I want to show you how the greatest growth companies routinely expand their TAM far beyond what investors initially expect.
And one popular stock whose opportunity is far bigger than investors currently estimate.
I first recommended Nvidia in April 2013.
Back then, it was a 28-cent stock (split adjusted), and still best known for one thing: graphics chips.
Gamers wanted better graphics. Nvidia made increasingly powerful GPUs. And the gaming market provided plenty of room to grow.
You could estimate how many gaming PCs would be sold each year, how much people would spend on graphics cards, and come up with a reasonable estimate for Nvidia’s opportunity.
But even then, Nvidia’s TAM was beginning to expand in ways few investors could fully appreciate.
In 2006, Nvidia launched CUDA, software that allowed programmers to use its GPUs for jobs beyond graphics. Researchers began using them for scientific computing. Data centers were adopting them to crunch enormous amounts of information. And GPUs were beginning to prove exceptionally useful for artificial intelligence.
A chip built to render video-game graphics was turning into a general-purpose computing engine.
The evidence started showing up in Nvidia’s numbers years before ChatGPT. Its Data Center revenue surged 145% in fiscal 2017 and another 133% in fiscal 2018. By fiscal 2022, Data Center was already a $10.6 billion business.
At that point, sizing Nvidia by the gaming market was like measuring an airport by counting the cars in its parking lot.
The opportunity had far outgrown gaming.
Then generative AI arrived around Thanksgiving of 2022 and sent demand into overdrive. Training and running AI models required enormous amounts of computing power, creating a market far larger than investors had anticipated.
About a year later, Nvidia’s Data Center business alone had grown to nearly twice the size of the entire company just two years earlier in terms of revenue.
As investors began to grasp just how much Nvidia’s TAM had expanded and it started showing up in numbers, the stock went on to surge roughly 13X.
Missing the lessons from Amazon and Shopify could cost you money too.
Investors have routinely underestimated both companies.
For years, investors saw Amazon mainly as an online retailer. The obvious opportunity was taking a bigger share of the money people spent in stores.
But Amazon kept expanding the market it could serve.
It opened its platform to third-party sellers, turning Amazon from a retailer into a marketplace. It built a massive logistics network. Then it began renting out the computing infrastructure it had built for itself, creating Amazon Web Services. Later, Amazon’s huge shopping audience created another opportunity: advertising.
Each new business made Amazon’s potential market much larger than investors had initially imagined.
Imagine trying to size Amazon’s opportunity in the late 1990s by estimating only how many books people would eventually buy online.
You would have missed almost everything that came next. Including Amazon stock soaring 200,000%.
Same story with Shopify.
Originally, investors could size Shopify’s opportunity by looking at how many merchants needed its subscription software to build and run an online store.
But that underestimated the real opportunity.
Those same merchants also needed payments, shipping, financing, fraud protection and dozens of other services to run their businesses.
Shopify began providing them. That meant its TAM expanded far beyond e-commerce software. And Shopify could earn more from the same merchant as that merchant grew.
Today, nearly 80% of Shopify’s revenue comes from these merchant services.
Shopify stock soared 30X in the last ten years.
This is the mistake I see again and again in our research.
Investors tend to treat a company’s TAM as if someone drew a fence around it. Nvidia sells gaming chips, so measure the gaming market. Amazon sells things online, so measure e-commerce. Shopify sells online-store software, so count the merchants willing to pay for it.
That works fine for an ordinary business. But growth companies have a habit of moving the fence.
As they grow, they accumulate assets and come up with new ways to solve problems that can take them into entirely new markets. Technology built for one product can solve problems elsewhere. A distribution network can carry new products. A trusted brand can open adjacent markets. Millions of existing customers can become ready-made buyers for the next service.
And the company doesn’t have to start from scratch each time. It carries those assets into the next opportunity, giving it more chances to find another large market.
Wall Street asks how big a company’s market is today. But the question to ask is where the assets it already owns could take it next. Because the best growth companies keep finding new markets to grow into.
We’re watching this happen again today.
Wall Street is underestimating Apple’s TAM.
Most investors still frame Apple around two giant engines: iPhones and Services. And with Apple unveiling its first foldable iPhone this month, the spotlight is naturally back on smartphones.
But in the background, Apple is building a much bigger opportunity that could push Macs into a market investors rarely associate with Apple today: AI infrastructure.
Its newest high-end Macs are becoming powerful enough to run large AI models directly on the device.
You see, not every AI workload needs to live in a giant data center.
Think of a lawyer running AI across confidential client files. An accountant processing tax records. A doctor working with sensitive patient information. Or a tiny startup using AI agents to automate work that once required a much larger team.
For customers like these, sending every document to someone else’s cloud can create problems around privacy, control, and cost. Running AI locally provides a better alternative.
And the potential customer base is enormous.
The US had 30.4 million businesses with no paid employees in 2023. That’s more than 80% of all US firms. Most of those firms will never buy a high-end Mac to run AI locally, but a fast-growing slice at the top will.
Apple already sells many of these businesses the computer. Now imagine that computer becoming part of their AI infrastructure too.
That would stretch the Mac opportunity far beyond the traditional PC market and into the massive spending boom surrounding AI. And that’s without spending a single dollar on datacenters.
Local AI is still young. The largest and most capable models still rely heavily on giant data centers.
But it’s worth watching. Wall Street is focused on counting iPhone sales. Wearables. And modelling Services revenue.
Most analyst forecasts still treat Mac as a mature PC franchise. They give little or no credit to how much local AI could expand the Mac’s TAM.
Related: The Growth-Stock Trap That Destroyed Zillow and Groupon
