As we covered last week, free services I Want Sandy and Stikkit are closing shortly, joining a growing list of Web 2.0 free-to-consumer startups that have shuttered their sites. It’s not just the little guys that are going out of business, either: Google Lively is set to become the latest failed experiment from the search behemoth later this year.
While just a few data points don’t make up a trend, it does seem likely that we haven’t seen the last closures. Services start up in a burst of optimism, then hit the cold hard wall of needing to pay for servers and bandwidth. The tightening of venture capital and the decline of online advertising have been covered elsewhere: other factors that will make it tough for free eternal-beta Web 2.0 startups to stay in business. But how is the savvy web worker to cope?
It’s clear that “free” doesn’t actually mean “free” when you put it into a larger context of web work. Choosing to use an online service for part of your business workflow carries with it opportunity costs and risk costs. Rather than rushing to put your entire business online, consider these guidelines:
1. Prefer in-house servers for mission-critical applications. Yes, it’s nice to be able to sit down in a cybercafe with instant access to all of your work – until the day when some of it is just missing in action thanks to factors beyond your control. As a software developer, for example, I’m not willing to place client code on a free source control server in the cloud, even though there are several excellent ones out there. Instead, I run my own server, and can access it remotely if I need to. There are tradeoffs, of course, but for this particular data I am willing to assume the administrative costs to eliminate the dangers of free.
2. Prefer portable data. If you do have data that you’re willing to host with a free service, then one of the best things you can do is ensure that you’re not locked in to that particular service in case things change. This implies that you should have easy access to all of your data in some standard format (XML or CSV, for example) in case you ever want to pick up and leave. Beware of services that have no export capabilities or that only let you save data in their own proprietary formats.
3. Prefer backed-up services. All the data portability in the world does you no good if you don’t have the data when you need it. I Want Sandy and Stikkit are doing the right thing and giving a few weeks of warning, but what if you end up on a service that shuts down overnight? We’ve mainly been concerned with backing up desktops into the cloud over the past few years, but I’m increasingly convinced that backing up the cloud to local storage is the sensible thing to do. I expect this to become a key differentiator in the next round of services.
In the current software and economic environment, I’m re-evaluating my dependence on free services, and pulling some things back to less trendy but more reliable client-side applications. How about you?