What happens to data when a Web start-up dies?
One of the greatest stumbling blocks that a start-up Web service company has to get over is customers' fear that the company will die and take their data with it. Manufacturers of traditional software can go belly-up without it immediately affecting a product's utility. And if General Motors went out of business tomorrow (I know, shocking), its cars would still be drivable. But Web services are different. When your cloud app goes under for the last time, it sinks customers, too.
In the best of the bad cases, when a Web service goes offline, the company in question is able to shut down in an orderly way and let its users know long before the servers are pulled off the Net. Large multiproduct companies generally do this well. When Yahoo shut down Yahoo Photos in favor of Flickr, it gave users plenty of time to offload their pictures, or move them to another service. When Hewlett-Packard shuttered the cloud backup system Upline, it likewise gave users fair warning.
But smaller companies may not be able to keep their servers up when their companies begin to fail. Does that mean you can't trust them with your data? Not necessarily. One company that's planning for the worst is Pogoplug, the makers of the cool, simple network file-sharing appliance I covered in April. Pogoplug's doomsday plan is not exactly fail-safe. The servers won't magically spring back to life if the company goes under. It's more of a seed vault kind of thing, but it's a good solution.
Pogoplug has put the source code for its servers in escrow. If the company goes bankrupt, the terms of the escrow dictate that the code be released to SourceForge as an open-source library. Then someone could come in and restart the service. Pogoplug CEO Dan Putterman says the service is designed to run on cloud services like Amazon EC2, so whoever wanted to re-light the system wouldn't need a large, expensive infrastructure to do so.
It's worth noting that Pogoplug doesn't actually store customers' data files. Rather, the online component of the product is a service that connects customers, their storage devices, and the users they invite together so they can share information. Were Pogoplug to go belly-up tomorrow, nobody would lose data, only the access to the transfer service they paid to use.
If the worst happens and Pogoplug does shut down, whoever restarts the service would have to re-gain the trust of users, who would need to re-configure their accounts and get new passwords to use the service again.
Even so, I like the dead-man's switch aspect to Pogoplug's escrow plan. While it won't guarantee that the product will resurrect if the company dies, it does at least make it possible--maybe even likely.
I'm surprised more companies, especially Web companies that store customer data, don't have public doomsday plans. If you buy life insurance, it's not an admission of weakness. It does not telegraph that you expect to die tomorrow. But if you've got a family (of customers) to support, the insurance can be a smart thing to have.
Rafe Needleman writes about start-ups, new technologies, and Web 2.0 products, as editor of CNET's Webware. E-mail Rafe. 



Their plan is like me going bankrupt but telling my creditors , too bad, that asset was promised to my kids.
Although if there is an existing "license" agreement with the users, any creditor or an acquirer would have to abide by the agreement. I guess you could do it with a license between them and their users that says that the code will be escrowed, I've seen this kind of arrangement in larger SW agreements so it could be legit.
www.storyofmylife.com and www.storyofmylifefoundation.org
Genieve, Eric, Mariebeth & Stephanie (gemsFamily)
http://www.my-msi.net - a plan for the rest of us!
- by netdocuments June 1, 2009 3:26 PM PDT
- as a SaaS ECM (content management) for the past 10 years, us small vendors must do the same as HP or Yahoo in providing a continguency plan for our customers....if we didn't have this we would have some of the national large law firms storing all their documents in our service. Redundancy is critical and exit strategies such as leaving the service up for six months following a bankruptcy and having that contractually established with the data centers. Having on-premise backup services that keep a copy of all data on the customers premises on an ongoing basis is also a strategy for customer peace of mind.
- Reply to this comment
-
(8 Comments)If any SaaS vendor is a serious vendor they must take these issues in consideration otherwise all these fly-by-night new vendors leverging Amazon EC2 or force.com better read the fine print.