[pg]

Actually there wasn't anything in DHH's talk that I really disagreed with, except the use of the word "startup" for the kind of company he was describing. There's already a word for companies like Italian restaurants: businesses. A startup is a very specific kind of business: one that starts small but could grow very large. Only a fraction of the 30 million businesses in the US are startups.

I think what made people think DHH's advice was relevant to startups was that he was talking about starting a business to write software, which is also what most startups do. But structurally the kind of small business he was talking about was more like a landscaping company or a shoe store or a restaurant. Which is a perfectly legitimate thing to do; it's just not a startup.

[dhh]

I think that's a pretty narrow definition of a startup. Also, there are plenty of businesses that start out small and then grow very large over time. Actually, most large businesses took a long time to get there.

Say like foot locker or Zappos for a "shoe store" (both big business) and say McDonald's or Olive Garden for a restaurant.

By the same measure, we're building 37signals to grow as well. Grow revenues, grow customers, grow influence. We're just not that hooked on growing head count or office space (which are often the most visible indicators of growth for a private company and thus often mistaken as the only indicators).

To me a startup simply means a new business that's getting off the ground. That business may well end up big one day and it may not, which is okay too.

I generally don't think that you can become a star by trying to be a star. I think you just try to be the best at what you do and if you are, hopefully the star part will take care of itself.

[pg]

I think that's a pretty narrow definition of a startup.

Maybe, but I didn't invent it. It's just the definition in current use. As most people use the word, it's more specific than a business that's merely new. As most people use the word, it doesn't include barber shops, gas stations, shoe repair stores, etc. There are structural differences between that kind of company and a startup. McDonald's and Olive Garden aren't restaurants, but restaurant chains. They do things very differently from your local Italian restaurant.

[dhh]

Most restaurant chains didn't start as chains, they started just as that single restaurant. Most big businesses didn't start big, they started small.

But that's again off topic. I do agree that the startup term is most frequently associated with technology companies, but you do hear about startups within a fair number of other lines of business too.

[pg]

Actually most chains are started as chains. Bizarrely enough I happen to know a lot about this, because the only reading material at the Chinese restaurant near my house in Cambridge is restaurant industry trade mags.

Sometimes the chain guys will take a restaurant that's successful and turn it into a chain, but when they do this they usually keep nothing more than the name and a few recipes. McDonald's was a burger place started by the McDonald brothers, but it only became the McDonald's we know after Ray Kroc bought it and transformed it into a chain.

I agree that startups don't have to be doing technology. I don't remember saying otherwise.

[tptacek]

This turns out to be easy to research; just start from the R&I 400 list of chains and go read their corporate histories. If you were in business for many years before opening up satellite locations, we can safely say you weren't started as a chain.

You're right; restaurants conceived as chains dominate the top of the list, with some very notable exceptions, including KFC, Starbucks, and Subway.

Something else that the top of that list shares in common is that they were lucrative for their investors.

Something else that they share is that they are all crappy. Look how far down that list you have to go before you hit something you'd actually choose freely? In-N-Out is #89.

[aswanson]

What then, qualifies as a startup per your definition, even in the technology space? You could say that Apple, Microsoft, and Yahoo started off along the local restaurant model, selling (or providing services to) a small group of local hobbyists and expanding naturally as new opportunities presented themselves. [EDIT: In each of these businesses (except yahoo, maybe) the goal was to grow a business. I don't even know if MSFT ever took VC. They just sold product until the IPO. I doubt they saw VC or a public exit as an upfront goal. They just ran the business to the point where it was a natural progression.

[pg]

I try just to use standard definition. A startup as most people use the word means a comparatively small, comparatively young company that is designed to grow very big. Most but not all such companies create technology, because technology scales so well.

Basically, the difference between a small business and a startup is like the difference between a shrub and a redwood seedling.

The metaphor isn't perfect, because occasionally companies are transformed from one to the other. But this is extraordinarily rare.

[menloparkbum]

Yahoo! became a business by accident. Apple was started by a guy who wanted to make the best personal computers, and Microsoft was started by a guy who wanted to put a computer in every home and on every desktop.

In the Apple/Microsoft initial period, there were plenty of smaller competitors who wanted to slowly sell stuff to hobbyists and "expand naturally" and none of those companies exist anymore.

[cellis]

I don't think Apple, Microsoft, Yahoo started off to build small businesses with family men. These were young , extremely ambitious, single people, out to change the world.

[mattmaroon]

You're absolutely correct. I have a friend who works for a company that does nothing but launch franchise changes, and they've had a few monster scores. They do it in a few ways. One is by, as you mentioned, taking a pre-existing successful restaurant and using the name and some of the menu.

Some times they will simply develop concepts (often 4 or 5 at a time) start one of each, and see what sticks.

I've studied the restaurant industry quite a bit and can only think of a few chains that stumbled into it.

[menloparkbum]

That may have been true back when McDonald's started, but most of the chain restaurants around today were started as franchises.

[thaumaturgy]

pg has been specifically keeping the definition of "startup" narrow, to describe only the businesses that go through the two-guys-plus-vc-plus-sleepless-nights-equals-fast-megabucks pattern that we saw in 2000. I think that's the model he's most interested in, that he sees as the most dynamic and rewarding.

Shiny.

That said, the things you describe closely match what I've wanted to pursue. For me, and many others, it's the more rewarding, long-term way of doing things. Your talk at SS08 was a huge breath of fresh air.

But, I don't think it's all that fair to dilute someone else's definition of a startup with your own idea of what a startup is. It might be time to come up with a different term, and let startup mean the thing that pg and the vc guys and others want it to mean.

[dhh]

I think that's a hijacking of the word that's not helpful at all. And I don't think that pg would argue that you can only call yourself a startup if you're going for VC funding and a sale or IPO.

If so, there's an awful lot of tech companies that have been wrongfully labeled as startups under this narrow definition. And you'd only be able to call yourself a startup in retrospect once you saw whether you ended up being a megabucks exit.

So 37signals would have been a startup if we sold to Google tomorrow, but not if we kept on as an independent company just making money?

[thaumaturgy]

No, he might not, but getting funding of some kind is a pretty large part of the kinds of startups that they /are/ talking about, because it makes it easier to develop that massive, rapid growth that gets people's attention. Likewise, a sale or IPO aren't the only ways to derive value out of the startup, but they're the ones that get the most focus, because they're the fastest way to get the desired results (lots of money). I think that this point of view was pretty clear in several of the talks at SS08.

Do I necessarily think that that's the only kind of "startup" that exists? No, not at all. (And I'm probably gonna get busted at some point for putting words in pg's mouth.)

But, in a nutshell, no, they're not talking about the same kind of businesses you are. Others have already made this point for me (http://www.gaborcselle.com/blog/2008/04/startup-school-surfi...).

So, coming back and saying, "No, we're talking about two different approaches to the same thing!" ... well, that's not helpful.

Again, my only point is that they're talking about a very specific sort of model when they use the word "startup". If you think your model is the same thing, I wonder how you'd distinguish between a startup and a "software business", or "web venture", or whatever. (And, I think this point is getting made by other elsewhere, not just in this thread but outside of news.yc.)

[tptacek]

According to Graham, "getting funded" isn't even a part of the plan for most YC startups.

[sabat]

So it's gotten this far now? DHH loves to start fights so much that he had to come over to PG's forum and start them?

PG, who is always civil here, dragged into a stupid semantic fight?

I boggle.

"So 37signals would have been a startup if we sold to Google tomorrow, but not if we kept on as an independent company just making money?"

No, man. You're just playing semantic games, and you know it . The company would have been a startup if its founders had the intentions of being large or becoming part of something large.

Really, you have nothing better to do than to come to this nice place and crap all over it?

[davidw]

I think it's quite interesting to see him here. He disagrees, in a polite way. That kind of thing is healthy. I don't really agree with dhh's definition of a startup, but there's probably not a sharp line between 'startup' and 'a new business' in some cases. In any case, I like the 37signals model, even though I have some nagging doubts about whether it's really quite so simple in an on line world where network effects and other economic factors are often very different from "the real world".

[sabat]

Polite?

You're giving him more credit than he's due. He is not polite. Passive-aggressive arrogance is not the same thing as "polite".

[pg]

pg has been specifically keeping the definition of "startup" narrow

Argh. I haven't been doing anything to the definition of the word. I'm just using it as everyone does. Think about newspaper headlines, for example. Wouldn't you be surprised if a company described in a headline as a "startup" turned out to be a shoe repair shop, or Exxon?

[brlewis]

Pardon me for quibbling: it's not about potential growth, but about having an extended period of time at the beginning without significant revenue. A shoe repair shop is not a startup because its debt is expected to stop deepening as soon as the doors are open, not because of its limited growth potential.

I got an accountant to do my taxes this year for the first time. He called my sole proprietorship a startup and did the accounting accordingly. Bootstrappers: keep records of all your expenses, even if it takes years for you to get everything going. Eventually it will all be deductible, amortized over 5 years from the time you start getting revenue.

[tptacek]

Would you be surprised to see Threadless described as a startup? What was American Apparel in 1997? A small business?

[Alex3917]

A startup exists to make extraordinary profits by arbitraging a temporary economic disequilibrium by doing more than anyone thought possible with less than anyone thought possible; the industry is irrelevant.

[tptacek]

I don't think that had anything whatsoever to do with my point. Is Threadless a startup?

[Xichekolas]

Am I the only one that thinks 'startup' simply means a new business without profits? This just seems like a silly semantic debate. 37signals was a startup once, it just happened to choose a different outcome.

Every startup should be free to choose their outcome, and if most choose to try and quickly flip, then more power to them. If they choose to try and build a profitable independent business, great! There is no need to turn the 'liquidity event' vs 'sustainable business' camps into warring factions, like some weird language flame war.

It was refreshing at SUS to see a contrasting opinion, but neither side is trying to argue that their way is the only way. Both ways obviously work, and I think anyone would be happy with either of them over not succeeding at all.

[menloparkbum]

One of my problems with DHH's talk was the examples he used. Starting your own restaurant is incredibly risky from a statistical standpoint. Even if you don't die in the restaurant business, you hardly make any money at all. And since this thread is ostensibly about work-life balance, the restaurant business is probably the only example DHH could have chosen where everyone works longer hours than a pg-defined startup.

Zappos was also a poor choice. It wasn't as if Zappos was a small time shoe store that decided to put up a web storefront and somehow magically scaled up its operations into a billion dollar business in the course of 5 years. Zappos was started buy a guy who had 250 million dollars from selling a software startup to Microsoft. It was specifically started to corner the online shoe market, in the same way Amazon was started to corner the online book market. It was also funded by Sequoia capital, further making it more like a "startup" startup than a DHH work-life-balance startup.

[sanj]

I think that this is an important point. The term "startup" has been narrowed to the point where it doesn't include what are disparagingly referred to as "lifestyle companies".

And I think that's really unfortunate because controlling the language controls the thinking.

Let's face it: people (at least in tech circles) think it is very cool to be in a startup. And it is in the VCs interests to further that sense -- of being caught up in something big and exciting.

It lends a rockstar sensibility to a startup that is NOT conferred to a lifestyle company.

Even reading at Paul's posting on this tends to reinforce that subtle dig -- look at the examples: "landscaping company or a shoe store or a restaurant". None of those sound exciting to me.

Perhaps my issue is that the measure of "grow very large" is ambiguous. Does it mean large profits? Sorry, Facebook. Large revenues? Sorry Youtube. Spending lots of money? Yay Webvan! Perhaps it means employing lots of people -- which the left-winger in me approves of mightily!

I've always felt that the goal should be to touch as many lives a possible. That's my personal metric on which I base "large". This is consistent with Paul's examples -- you can touch a limited number of people with a landscaping company or a shoe store or a restaurant.

But on the web you can touch a LOT of lives with a small set of money. I would hazard that David's touched far more lives at 37Signals through their consumer products and through Rails than all by one or two of the startups on YC or part of YC.

That's large.

In my world, that counts as a startup.

[cstejerean]

how is the business model of wufoo different from 37 signals? Are they a startup or a shoe store? A startup should simply be a company that is just getting started. Once they become profitable feel free to call the a "business".

[r7000]

A company that starts small but grows very large could describe exactly the 'Italian Restaurant' sort of startup. On the web you have a much bigger neighbourhood than any Italian restaurant. The limitation on a landscaping company, for example, is its ability to scale and the geographic area it can serve.

[tptacek]

I think Graham's analogy is more apt that he realizes. Here's where it takes us:

Yes, the majority of breakout chains/startups are started as chains/startups.

But, plenty of breakout businesses aren't built to scale out fast. Starbucks didn't break out for over a decade. Microsoft didn't get the DOS contract until '81.

Meanwhile, lots of small businesses are small deliberately. Thomas Keller isn't franchising Per Se. Grant Achatz isn't franchising Alinea.

So what does it tell us when a company runs a single restaurant for 10 years, or refuses to take VC funding or hire ahead of revenue?

Exactly nothing.

[menloparkbum]

Celebrity chef restaurants are not really businesses - they are more like old-school arts patronages. They are started with funding by rich finance guys who are looking for a hobby project to make themselves feel cooler. The only people who go there are also uber rich finance guys, other people in the celebrity class, and the occasional not-as-rich foodie. As soon as the rich people and celebrities get bored with the restaurant, it shuts down.

Alinea was funded by 8 futures traders who each put in half a million dollars.

[tptacek]

Also: be more right. =)

From the site:

Opened in May 2005, Mr. Kokonas says Alinea grossed $3.5 million its first year, an average of $14,000 for every night it was open. There are eight investors: The smallest stake was $75,000, the largest Mr. Kokonas', at just less than $500,000. Return on investment is expected to exceed 20% per year, he says.

This doesn't sound like an old-school arts patronage to me.

[mtts]

Not that I'm an expert on the celebrity chef restaurant business or anything, but I can name one example where this is at least partly true: Ferran Adria's El Bulli (voted best restaurant in the world a few years ago) actually operates at a loss. Adria supplements his income by putting out books and by selling some of the more unusual ingredients he uses.

(edit: typo)

[tptacek]

Charlie Trotter, Frontera, French Laundry, Babbo. All successful businesses. That there are unsuccessful businesses in the same category does not make them "old-school arts patronages".

[menloparkbum]

The economics of celebrity chef restaurants are so different from anything else that they resemble arts patronages more than they resemble the economics of other businesses, including other restaurants.

You have to already be famous to start one, you have to have met a rich guy who thinks investing in you will make him cool, you only cater to a class of people who in previous centuries would be known as "the aristocracy." Etc.

[tptacek]

Still wrong: Keller wasn't famous before French Laundry, Bayless is famous because of Frontera. Support claims with evidence.