The Wide World of Web Apps
Well, as I mentioned before, my next application will be a web app. Web applications have kind of a bad rap around the mISV community (which seems to mostly be desktop app folks). This could be because people like Paul Graham have recently been saying that Microsoft is dead because there is no need for Windows applications anymore. I don't think there needs to be any bad blood between desktop and web app developers at this point. While many many things lend themselves better as web apps, some things just make no sense to be anything besides a desktop app. I think the future is a combination of the two (online storage with offline functionality).
I think another reason web apps get such a bum rap is that all of the "popular" ones are things that have no clear profit angle (besides advertising), such as social networking sites. Or there are apps like Basecamp that just come off as too simple. But most of the really profitable web applications don't fall under either category, and are decidedly unsexy.
1) Consumer web app - Examples: Flickr, MySpace, FitDay, BudgetSimple. When people say there is no money in web apps, this is what they are thinking of. 90% of Web 2.0 falls into this category. The applications can be used by normal people, often have no cost associated with them, and sometimes have advertising. I really do believe that people are not yet ready to buy online applications. For example, Andy has some Web 2.0 wedding planner competition. Google recently has came out with spreadsheet and word processing software. Sure, these offer some benefits over traditional desktop software, but none works as well as the desktop versions.
And it doesn't seem your average joe is ready to "buy" software they can't hold in their hands. If you make one of these types of web apps, get ready for lots of competition, count on a revenue stream that is entirely dependent on advertising, and pray you can launch TechCrunch and 37Signals type publicity.
Pros: When these go well, they go VERY well ($$$$ buyouts)
Cons: Like winning the lottery, quality doesn't matter as much as marketing and luck. In the meantime, you are not making money.
2) SaaS - Software as a service. Examples: Basecamp and Salesforce. They offer many benefits, such as always being up to date, instantaneous bug fixes, and (presumably) guaranteed data protection. Instead of being tied to a laptop, you can access your application and data anywhere. While the revenue stream is clear here, businesses are the primary people who would use these, and I think it's limited to only small businesses at this point in time. Large companies want internally hosted applications, and don't trust others for sensitive data, and if the application is down, heads will roll.
Pros: Customers must purchase your software every single month, creating a reliable revenue stream. Fixing bugs is easy as everyone gets the latest version.
Cons: You will not sleep, wondering if the servers are up. Selling these to consumers is difficult, selling them to corporations also may take some work.
3) Shrinkwrap web apps - Examples: Fogbugs, HelpSpot. This is the least popular arena of web apps, which is a head scratcher. These types of applications have all the benefits of a hosted web app, plus all the performance benefits of a desktop application, without complicated deployment schemes. I have seen completely unknown web apps like this sell for $300k per customer, with $100k a year support contracts. If you ever think web apps can't be profitable, email me and I'll tell you about the horribly designed web apps I see making a mint in this category.
Pros: Ease of deployment. Corporations love not having to install software on every single desktop.
Cons: Similar negatives to Desktop apps in terms of debugging problems, keeping people up to date, etc. Language and platform matters a lot more here.
So which one am I building? Based on my descriptions above, you would think it's certainly not #1. But you would be wrong. I'm actually working on two projects right now. The one I am really enjoying is the consumer app that has sketchy revenue prospects. It's an idea that I think is really needed, yet surprisingly does not exist. There are some clear ways to monetize it, so it's not quite as much of a long shot as MySpace.
The other project I am working on falls under category number 3. There is existing competition in this arena, but there is literally only one competitor, and their product is horrible. But because they hold a monopoly, they feel no need to improve their program. This is in an area I know very well, and this is the more likely of the two projects to actually make money.