I asked tonight on Twitter for web app ideas and got a bunch, some of them very interesting. But, if you told me your idea and I seemed skeptical, it's because I have a set of rules I'm operating against. I don't define them as The Right Way; they're the operating boundaries of the current experiment, and the experiment began in December or January (and will run about a year). If ideas don't fit within the operating boundaries of the current experiment, I automatically discard them without regard to merit.
Boundaries include: quick to build; cash-flow positive within a month, preferably much sooner; low maintenance; low overhead (e.g., nothing which requires forming a corporation). There are probably others; these are the ones I'm aware of. I haven't asked for app ideas before so I haven't needed to name my experimental constraints explicitly. I'm articulating something I haven't articulated before, and it's a little like transcribing the melody of background music that you've only just noticed a moment ago.
Of the constraints I've listed, the requirement to become cash-flow positive almost immediately is probably the killer. If you had no requirement to become cash-flow positive ever, you could easily come up with many useful web apps which required low maintenance and low overhead, and were quick to build. I set a goal to create one of these every month in 2009, and every year, the Rails Rumble churns out a bunch of apps like this in a single weekend. Giving yourself six months to cash-flow positive would expand the range of options considerably; that's even true for just three months.
However, the fundamental assumption behind this experiment is that venture capital operates on severely mistaken assumptions of cost and time to market, based on the economics of a previous era in technology; and that open source software and commodity hardware have made modern web development so agile, lightweight, and inherently cheap that new, robust, and ultra-tiny categories of business will result. Startups can work with "runway" periods of a year, or three, or five; but what if startups don't need any runway at all any more? People say that to launch a jumbo jet requires a long runway, but that's not true for a Harrier jet or a helicopter.
Or a freakin alien mothership
It's not true for a bumblebee either, and there's tiresome hubris in the jumbo jet analogy. Jets are very reliable, but they're not at all fault-tolerant - their failures are few, but utterly catastrophic - while swarms of insects form some of the most fault-tolerant systems in the universe. Open source software comes from a place of sharing in abundance, of combined collaboration and competition - what you could call "collabpetition" - and this is fundamentally different from the economic thinking of venture capital, which revolves around leveraging gigantor investments into even-more-gigantor wealth. I'm not opposed to gigantor wealth at all - my desktop background is a picture of a Murcielago - but the world's seen some very serious catastrophic failures of giant systems for turning huge wealth into huger wealth recently, so I'm very interested to discover what the bumblebee end of the spectrum is going to look like, exactly.
vroooooooooom! btw, I stole the "collabpetition" thing from Matt Knox
This is part of what's fueled my interest in internet marketing; IM businesses are tiny. It's usually one person, one product, and one web page written in HTML. A sophisticated IM web app adds a blog, a few more static pages, login functionality, automated e-mail list management (usually via Aweber), ad management with Google AdWords and Analytics, and/or a few other bits and bobs. Even though there are many parts, none of them are very groundbreaking to code; you won't need support vector machines, and your code may not contain the word
lambdaanywhere. But that simplicity comes from success. Lots of people have made lots of money with these business models, and already, this dumb beast, having proven itself a survivor in its niche, is wandering into adjacent ecosystems.
All the same, it's still early days. What evolution holds for these businesses is still unknown, and the future shape of the bumblebee may transform.