Also, while I agree that there's effectively infinite demand for new software (I can't imagine a day where all software companies and open source projects stop writing code), it doesn't mean that there's infinite room for new software businessnes -- especially the ones that start with a few hackers in an apartment cranking out a webapp in 2-4 months.

Can't answer directly for pg but I know my whole field (electrical engineering) has pretty much been converted to software. Musical instruments, medical diagnostics, anything that deals with information processing, which in this day and age, seems to be migrating to a software function.

But like I said, can't speak for pg.

If you aren't going to pay for a solution to a need, then it isn't a need. If you can't afford it, that is a technological issue. There is no such thing as someone with "too much money".

[edit: there are certainly things that money can't buy. Though I'm sure you can mail-order a lovely bride, a Zen Buddhist sensei fresh from Japan, and adopt some children in need to give your life meaning]

I say: you can have a need, but you can have not enough money to pay for a solution to that need. For example, you can say that as there is so much poverty, there is a big need for food so there is a big market for food right? But probably, the sector that is suffering poverty wouldn't make an interesting market, as would not have money to pay for the food you might offer to it.

That's an exaggeration, but the same can happen with other stuff. There are lots of human needs you can point to, but the segments that would pay for solutions to these needs are, as I think, more limited, for many reasons. Maybe your solution is expensive for them, maybe they don't want to use credit cards online, etc...

About "too much money"... that was where my "non-native" English came along. I meant: maybe the people that have the needs have not enough resources (money) to buy your product/service, making that an uninteresting market despite being an unattended need. I was not saying you can have "too much money".

You: if people don't know how to make something cheaply enough it's not a market.

Ivan: it could be a market if you solve the technological problem of making it cheaper.

You: making food cheaper isn't a market because the poor have no money.

Ivan: making food cheaper isn't a feasible market because nobody knows how to make it happen.

There, now you can return to the interesting question, continue arguing about what fraction of the needs you can think up are technologically feasible :)

I can think of all kind of needs: a google maps with 2d shading would make it fun for kids (to say something stupid). But they wouldn't pay for it, it would be very expensive to do unless you are google, and it would probably be expensive _even_ if you are google (because the raw data are photos and no 3d info excepting some selected places), etc etc etc. That's not an interesting need if you look from the market perspective... but if you fall in the trap "anything useful for anybody is worth doing" you might fall in the trap of creating that kind of worthless project.

This reminds me of Joel Spolsky's take on the 80/20 bloatware myth. The idea being that bloated programs like Word are bad because 80% of the users only use 20% of the features. But they don't all use the same 20% is the problem, so you can't just remove 80% of the features without pissing off a sizable group of people. I think the same applies to web apps. Nobody needs every one of them, but the right set of web apps for any given person probably isn't identical to anyone else's.

Read Seth Godin's book Permission Marketing if you want to learn about market creation and demand-driven design.

Maybe your ideas aren't specific enough.

Take my own not-yet-launched project (http://www.tarsnap.com), for example: Online backups. Lots of people have done those, right? Mozy, Carbonite, iDrive... google can easily find dozens of them.

But that's just the "30 thousand feet" view of what I'm doing. I'm not just doing online backups -- I'm doing secure online backups. Secure in the sense that I can't steal your data. Secure in the sense that the NSA can't steal your data.

Have other companies done online backup? Absolutely. Can anyone else reasonably describe what they're doing as "backups for the truly paranoid"? No -- that's something which nobody else has done.

I would suggest starting with a secure chat system with good logging options. That way you can reasonably target teenagers, bankers and the truly paranoid. And you don't need a huge backup infrastructure or a 50$ / year price tag to compete.

As I said, lots of people are doing online backups. Nobody else is doing secure online backups.

Building a secure system is a hard problem

Exactly -- that's why it's something which I can do but nobody else has done yet.

and [...] few people can judge how well your (sic) doing

Fortunately, with a doctorate in Computer Science, as the FreeBSD Security Officer, and as "that guy who found a security flaw in Intel processors a few years ago", I'm ideally positioned to show people what I'm doing right -- and what everyone else is doing wrong.

Is secure online backups a smaller market? Absolutely -- most people don't care about security. That's the price you pay for entering an arena where there are many established companies: The best strategy is to carve off a smaller part of the arena rather than trying to fight for the complete arena.

In the end security is more about trust than the algorithm. I need to trust you not to log the password somewhere etc. Established players gain some trust because they have other things on their mind than your data and presumably they want to protect their reputation, so IMO the new guy is bester off focusing on something more tangible so you get a chance to build your reputation.

Which is why I suggest a free App that demonstrates you do security. You could build that relationship of trust and drum up some somewhat free press around how subversive your application was and turn that into a strong market.

PS: Good luck with your startup. I think you can make it work, but after considering the same idea and talking to people about what they trust I moved on to other areas .

Feel free to send me an email (address on website) to let me know what you'd like tarsnap to do -- the more feedback I can get, the better tarsnap will be. :-)

I don't think I'm going out on a limb if I say that I think changes will come much more rapidly in the future than they have in the past. So, while the telephone has been around for over 100 years, I don't think Skype will last out the next seven years. Facebook will be lucky if it gets five years at the top (if it ever bests MySpace and makes it to the top). I don't think a better "social network" will dethrone Facebook...I think a whole other class of product will take its place (actually a dozen or more classes of products, probably, since people do so many different things with Facebook). Your job, as a technology innovator is to spot those trends and build the products that enable them.

And keep in mind that little web applications aren't the only problem worth solving. There's a lot of software that runs inside the firewall at businesses--and a lot of it is moving to web-based variants. There are many worlds to change, not just the world Facebook is addressing.

That said, there aren't many opportunities to make something as big and world-altering as Google. There's only so many problems on the web that touch every single human being that uses the web (that's a big customer-base). Search is actually the only one I can think of. Email, perhaps.

I can think of many ways to make things 10% better, but that's just the next set of features these guys need to implement. I'm not convinced that is a business.

When Google started to work, it was definitely 10 times better than the competition, or had redefined it.

Don't make me think about my kitchen! it's awful! I can't seem to keep it clean and the traditional solution to this problem, marriage, has apparently run its course. My girlfriend is a software geek as well and hates to clean the kitchen (and vacuuming) even more than I do. So where is that long promised robot revolution? Bring it on!

And where is that website that lets me auction my food orders, to be delivered by the best bidder? And knowing about the food I eat, the site could also tell me which nutrients I'm supposed to consume more of and which ready meals contain them, excluding the ones I hate.

There are so many old unsolved problems. Anonymous hassle free micropayments for instance...

Someone please make this, much lighterweight than PayPal, and saturate the market with it. Add a prepaid micropayment card anyone can buy at a store with cash.

This is one of the things that I have great interest in using (both as a user and as a client), but not so interested in building and running it myself! :-/

The reason why nobody does it that way is probably that it's difficult to garantee settlement and the credit risk involved. But do we really need such a garantee in cases where no physical goods are delivered? Maybe it's enough if it works in 90 % of all cases. That's more than the rate of legally acquired software licenses today.

In an attention economy you're fighting for eyeballs. And there's a smaller group that's always looking for what's new and different. If you want to pursue this demographic, and the larger, lagging crowd behind them, then you have a chance at capturing attention.

I share many of your concerns, and personally use very few Web applications--basically maps and mail. I often ask the same question: what can I build?

Perhaps it's better to choose a pre-existing category/market and try to win through better execution. I also think success is more likely when building software as a product instead of a free service.

Make a list of 10 absolutely fabulous web products that you love. Now draw a line through all of them that came to market as the first mover. I'll bet you don't draw a single line.

As long as businesses want more money and want to save more time, there is always room for b2b products that can demonstrate good ROI.

On the consumer side, there's certainly plenty of room. I just read that Meebo is reaching 30 million people per month. Twitter is functionally brand new and kicking ass.

The good news is that most consumer sites transition from a "providing value" stance (in the early days) to an "extracting value" stance (in the later days). Later stage consumer sites add more advertising, do more biz dev deals, and overall muddy up their value prop. Meanwhile, the next round of consumer startups leap into the fray with a nimble team, the next rev of web technologies, and a desire to provide value and build something that's better than the big guys. Compuserve, Prodigy, AOL, Friendster, MySpace, Facebook... They are/were all on the same road.

Will there always be new opportunities? Of course. At the moment, there are also still lots of opportunities for web applications.

Businesses on the other hand (rational ones anyway) will buy anything that exceeds a certain ROI, provided it meets certain requirements. There are so many ways to do this, but they aren't technologically sexy in the way consumer startups are. For instance, find a way to improve measuring their ROI! Then they're not guessing, you can point to exactly how much money they'll make/save. Improve something by 15%, then multiply that savings times the thousands of corporate users for one buyer. Viola, an upgrade license. Joel Spolsky is a master of this: see http://www.joelonsoftware.com/articles/fog0000000052.html There will always be more of a market for things like this.

On the XIX century people believed that everything has been invented.

Why the net should be different? It is still on a very early stage, we have only seen what it can provide. I think that there is plenty of room for new web startups!

They didn't have microwave ovens in 1987 ?!

"20 years ago" is not a fixed date ;-)

If you go to any lesser-developed countries they are not entrenched in "google" or "facebook" or anything for that matter because they don't have computers --but they do have mobile phones.

Everybody's talking about China & India, great. But what about the many nations on the African Continent? In five to ten years, countries like Nigeria, Kenya, Ghana, etc. will be the new technological and economical hotbeds.

Retail products are totally different than software as service which is free. Thus the barrier for the user to try new stuff on the internet is zero comparing to buying something.

Also, the barrier to create automobile technology it's obvious very high and discouraging. However, enough costly innovation is occuring from the existing companies. Hubrid cars are becoming reality. A model car from sketch to manufacturing to get in the roads is at least a 5 year effort. Compare that to developing an app in 2 months!

Any time you encounter pain or desire (yours or somebody else's) that's information-related, you've stumbled upon the opportunity for a web and/or mobile startup. When you meet that need, people's demand for time online will grow to encompass what you've given them.

As long as the kitchen isn't your Universe... :)

You answered your own question. What this means is that if you are serious about your start-up, you should shoot for nothing less than being in the 5-10 "really useful" list.

It is not easy by any means and pg has said before that as more and more start-ups launch, the difference between great ones and average ones would become more apparent.

I don't think this number is due to a lack of web applications to fit people's needs, but limitations of the wetware itself. This implies that there is a biological driver for web application consolidation. It also implies that whatever your app, you're already competing for "web app" time in the user's mind with some other site, even if it is completely unrelated. Futhermore, it implies that great non-consolidated apps actually cause some harm, assuming they take people away from other sites that are providing value to them.

Pure speculation on my part, of course. I just know that while I've seen a boatload of crazy-cool web applications, the number I use is fairly static -- I just shuffle some in and out every now and then.

Great question!

There is no limit to our use for knowledge. No point at which things could never be improved again. No end to progress.