Online Service Competition, Lock-In, Migration

There are many areas now where online services are competing for my business.

Blogging, for example (LiveJournal, Blogger, MSN Spaces, etc), photo hosting (Flickr, ZIMoBLOG, TextAmerica, PhotoBucket, etc),  music downloading (Napster, MSN Music, iTunes, etc), email (gmail, Hotmail, Yahoo Mail, etc).

Companies competing for my business is really nothing new, but ease of data migration is a relatively new issue.  

Most software companies realized a few years ago that users want access to their data.  In the old days, every application had its own binary data format, and you were stuck with it.  Companies viewed this as a way of keeping hold of their customers – the customers had too much invested in the data that only their application could read for them to migrate.

Then there were the days when everyone tried to figure out everyone else’s binary format.

Now, most applications have a way to export to some generic format (be it XML, or even a text file), and most applications can import.  If the export doesn’t match the import you can usually massage the data to make it work.  

Online services seem to be starting this trend over.  With the exception of music downloading, these websites are online applications that are letting me work with my data using their servers.

Most of them provide no way of exporting the data.  Gmail, for example, lets me import contacts, but not export them.  There’s usually no way to export your blog entries, or the metadata that you create for your photos on the photo hosting sites.  

There are a couple of ways this could go.  Here’s the one I think would be most interesting.

Create an abstraction layer.  Define an interface that these services can use to store my data.  Then let me choose where to point that service.

For example, lets say we’re using WebDAV.  I’d like to point Blogger at my server and say “put my blog data here”.  Later, if I want to switch to MSN Spaces, I want to tell MSN Spaces “my blog data is here” with the same URL, and have it pick up my Blogger data.

This could be almost invisible to the user – the default would be to use a store local to the service.  

Some advantages to this:

·        If I want to host my own data, I can.  Most people wouldn’t, but if I wanted to have gigabytes of data, I can because it’s my disk.

·        If the front end provider (say, Flickr) goes away, my data doesn’t go away with it.

I think something like this will inevitably happen; it’s just a matter of how long it will take.  Right now everyone’s still in “lock-in” mode.