I think a lot of "startups" would be better off calling themselves projects. There's nothing wrong with being a project: Wikiepdia's a project, GNU/Linux is a project. But neither of them ever expected to make money; the same could be said for many (by no means all) web apps that call themselves startups.

if it is to make money, it is a startup. if it an application on the web, its a web app.

it can be any combination of true or false for either.

And web apps can switch state. Delicious wasn't originally a company, but it turned into one.

I've been reading Founders at Work and I've been thinking about this. It seems that there's basically two main paradigms when it comes to tech startups:

1) Start out building a product/service that you think people will pay for (Viaweb, Excite, VisiCorp) and evolve the idea where appropriate

2) Start out building a product/service that you want to use, and its immense popularity makes it become a company (Facebook, Delicious, Yahoo, Google)

A lot of people (myself included) have thought "the way to make it big is by doing #2: making a neat website, and hoping it takes off incredibly". But perhaps a more statistically sound approach is to do #1. Has anyone done a study of this kind of thing and looked at expected value?

These two are not exclusive. The first computer Apple made was something the founders wanted, but also something they knew people would pay for.

Either way, you have to make something people want. I'm reading Four Steps to the Epiphany now...hopefully it will successfully explain how to figure it out what it is that people want.

The startup is the business itself. A web app is an app on the web. Gmail is a web app, even though it's a product of a largish corporation. Windows Live is a collection of web apps made by a company that hasn't been a startup since the Altair was the hot new thing.

Likewise, a startup doesn't have to be or use a web app. Loopt runs on cell phones.

Arguing about the definitions of words is productive if you're a lexicographer or a lawyer.

For the rest of us, it absorbs time which could be used to separate customers from their money by selling them things they want.

Well, it is good to be clear about what things are and where you'd like to go with them. The 'project' comment elsewhere here is right on - not everything can or should be about making money, and it's good to recognize what sort of thing you're dealing with at any given point in time.

You are obviously a startup then ;)

I think the webapp/startup line blurred together in the funding frenzy of the past few years where revenue-less webapps could be acquired profitably for the founder.

But the line is growing more distinct starting as revenue regains its importance in the scene. Startups are projects with clear revenue goals, and relatively detailed plans by which to attain those goals, that the founders expect to turn into some sort of business.

Webapps may just be useful and fun distractions without concrete monetization plans and/or goals. They have the potential to grow into startups, but are missing a lot of the other components that go along with it (business planning, revenue projects, goal deadlines, etc.).

That said, there's still no absolutely clear delineation. Some may alternate between the two definitions.

I always considered a startup as an incorporated legal entity and a web app as an application on the web.

At what point do you incorporate?

When you absolutely must, ie. something would break horribly if you weren't an incorporated legal entity. Like you have a term sheet from investors, or you want to start charging people. Incorporating early is fairly pointless, especially if you're just at the idea stage.

So you need a formal business structure to charge anyone any amount of money? I just made a proposal to advertise for a local business that will get me two figures per month if it's accepted, covering 25-50% of my current costs. I would rather not drop the money on an LLC until I have some idea of how it might get recouped...

Incorporating gives you better liability protection if you do it right. Depending on your product, this may be valuable.

Even if you're not incorporating, it's still a good idea to have a fairly formal agreement if you're working with a team of people.

To me a startup is something which is built with hopes for it to eventually be profitable. A web app is something that you build that you don't really expect to make money from

A web app is a product. A startup is the structure that monetizes/wants to monetize the web app.

They aren't mutually exclusive.

"A startup is a human institution designed to deliver a new product or service under conditions of extreme uncertainty." -Eric Ries

They're independent from each other. A startup is young company/business offering a product or a service. A web app can be the service that a startup provides, which is why you see many web startups as web apps.

I can create a web app without making it a startup/business, and vice-versa.

A web app is a product. A startup is a business. A business can field one or more products or services in an attempt to make money. In the case when a startup has one main product, a web app, the terms are sometimes conflated (which I don't really like).

A startup offers services or products, some of which could be web apps.

A web app is another form of software distribution, via the web.

Just to piggy bank onto your comment:

A start-up doesn't necessarily have to be a webapp or even in the technology space. I believe any new business that could disrupt the status quo in an industry to the benefit of customers could be called a startup. In it's early days, Dell was considered a startup because of its disruptive distribution system, not necessarily the technology.

I think the question should be more precise. What's an _internet_ startup and what's a webapp?