What is a startup? A plain-English definition with examples

Your cousin opens a bakery in Turin. Your flatmate spends his evenings in Milan building an app that matches students with spare rooms. Both have started a company, both are taking a risk, and both will work far too many hours. Only one of them has started a startup. The difference has surprisingly little to do with technology.

So what is a startup, exactly? In about seven minutes you’ll have the two definitions people in the industry quote, the three words that separate a startup from a small business, why the Italian label “startup innovativa” is a different thing, and a short test for your own idea. Think of it as the first stop on our beginner’s map from zero to a first round.

What is a startup? The two definitions to remember

The definition you’ll see quoted most often comes from Steve Blank. In January 2010 he wrote that a startup is “an organization formed to search for a repeatable and scalable business model.” Two years later, in a post called Search versus Execute, he added one word. Temporary.

That word carries a lot. The search is meant to end. Either you find a way of making money that works again and again (repeatable) and keeps working as you get bigger (scalable), and you become a normal company that executes, or the money and the patience run out first.

The other definition you’ll hear is Paul Graham’s, from a September 2012 essay titled Startup = Growth: “A startup is a company designed to grow fast.” Being newly founded isn’t enough, he wrote. Working on technology isn’t required, and neither is taking venture funding. “The only essential thing is growth.”

He added a line that clears up a lot of confusion: “Millions of companies are started every year in the US. Only a tiny fraction are startups.” Italy’s chambers of commerce registered 323,533 new businesses in 2025. Every one of them is a new company. That on its own doesn’t make any of them a startup.

Growth, scalability, uncertainty: the three words that matter

Growth. Graham measures it by the week. During Y Combinator’s programme, he wrote, a good rate is 5 to 7% a week, 10% is exceptional, and 1% suggests you haven’t yet figured out what you’re doing. That sounds tiny until you compound it. Do the maths. Grow 1% a week and you’re 1.7 times bigger after a year. Grow 5% a week and you’re 12.6 times bigger.

Scalability. Here’s the test: does serving your next thousand customers cost far less than serving the first thousand did? If yes, the business scales. A bakery can’t really do that, because more bread means more flour, more ovens and more people awake at four in the morning. Software, platforms and marketplaces can. Airbnb began in 2007, when two hosts put up three guests in their San Francisco home. Today it counts more than 5.5 million hosts and over 2.5 billion guest arrivals. It owns none of the homes.

Uncertainty. This is Blank’s “search”. The baker knows people buy bread, roughly at what price and at what time of day. A founder is guessing on almost everything: who the customer is (a company or a consumer, what people call B2B or B2C), what they’ll pay, how you’ll reach them, whether they’ll come back. Your first months are about turning those guesses into facts.

Miss one of the three and you probably have a small business. A good one, maybe.

Startup vs small business: what’s the difference?

Picture a bar in Milan where people pay for their espresso with Satispay. The bar is a small business: a proven model, local customers, profit (with luck) from the first year. The app on the customer’s phone is the startup.

Satispay was founded in 2013 by three people from Cuneo, Dario Brignone, Samuele Pinta and Alberto Dalmasso. They had the idea in 2012 while still working other jobs, then quit to work on it full time. The first €400,000 came from sixty people who believed in them. By June 2017 the app had 130,000 customers and 18,000 shops. Today the company says it has 6.5 million users and more than 750 people.

	Small business	Startup
Goal	Steady profit and a good living for the owners	Fast growth toward a very large company
Growth	Gradual, often local	Fast, aimed at a big and often international market
Business model	Known and proven	Still being searched for and tested
Funding	Savings, bank loans, profits	Shares sold to angels and venture funds, sometimes grants
Risk	Moderate and easy to see	High: the model may never work
Exit	Often none: the owners keep the business or hand it on	A sale or a stock market listing, which is what investors wait for

None of this is a ranking. Plenty of bakeries make their owners a better living than the average startup makes its founders. They’re simply different games, played with different money and very different odds.

Is a “startup innovativa” the same thing?

No, and in Italy people mix the two up all the time. The startup innovativa is a legal status created by Decreto-legge 179/2012. For the special section of the Registro delle Imprese, the company has to be young, set up as a capital company (an SRL, for example), and spend most of its energy on innovative, high-tech products or services. On top of that it must clear one of three hurdles: R&D spending of at least 15% of the larger of production costs and production value; a team in which at least a third are PhDs or researchers (or two thirds hold a master’s degree); or a patent or registered software in its name.

Since the 2024 annual competition law (Legge 193/2024), keeping the status beyond the third year means hitting specific targets. The benefits are real, and our guide to Italy’s innovative startup status and its requirements walks through them. And yes, a company can shoot upward without ever registering, while a registered startup innovativa may run, day to day, like any small business.

How are startups funded?

A bakery can walk into a bank, borrow, and pay the loan back out of its profits. A startup usually can’t. There are no profits for years and very little a bank would accept as security. So it sells pieces of itself: investors buy equity, a share of the company, knowing that many of their bets will lose money and hoping one big win pays for the rest.

The order tends to repeat. It starts with the founders’ own time and savings. Next comes money from friends and family, then individual investors known as business angels, and finally venture capital funds, which invest other people’s money. Each step is a round with a name and a typical size. Satispay’s sixty early backers were followed by a first €5 million round, half of it from Iccrea.

Investors get paid mainly at the exit, when the company is sold or lists on a stock exchange. That’s why they ask about market size and growth long before they ask about profit, and why even a wonderful, profitable bakery is not a venture investment. The next step on the map explains how startup funding stages work, from bootstrapping to Series A.

Is my idea a startup? Six questions to answer honestly

Get a “no” and nothing is lost: you’ve just learned which rules you’re playing by. If a word trips you up, our startup glossary for beginners explains fifty of them in plain English.

1. Say demand grew tenfold tomorrow: could you serve it without hiring ten times the people or paying ten times the costs?
2. Is the market big enough that the idea could work well beyond your own city and country?
3. Can you not yet say who pays, how much, and how you’d reach them? That’s the search Blank describes.
4. Would you give up part of the company to investors in exchange for speed?
5. Can you live with a few years of losses if the growth is there?
6. Is there a reason this can work now and couldn’t five years ago, such as a new technology, a new rule or a new habit?

Mostly yes? Then the next question is whether anyone will actually pay, and our seven-day plan to validate a startup idea without spending money shows you how to find out. If cash is the worry, read how to start a startup with no money. And for proof that big companies start tiny, here are Italian startups that started small, plus our list of Italian unicorns for the few that went all the way.