Social networks operate on the premise that it should be as easy as possible for a user to input their data into the system: user profiles, blog posts, media etc, but if they try to move their data someplace else or simply want to make a backup, more than often they're out of luck. After all, why would any company make it easy for its customers to leave? It's for this reason that amongst web 2.0 evangelists, data portability is being promoted as a much needed antidote to customer lock-in.
Speaking at the recent Web 2.0 summit, Google CEO Eric Schmidt was quoted as saying:
If you look at the historical large company behavior, they ultimately do things to protect their business practices or monopoly or what have you, against the choice of the users... The more we can, for example, let users move their data around, never trap the data of an end user, let them move it if they don't like us, the better.
Marshal Kirkpatrick (in an earlier post on TechCrunch) went as far as to call data portability a rights issue:
I own rights on my data; I want to be able to easily and quickly take it with me from one social network to another.
But outside of the web 2.0 crowd, do users of social software really care about such technicalities as data export? And why should companies support data portability? I put these questions (along with a few others) to four leading developers in the social networking space.
Marc Canter (CEO of Broadband Mechanics): "Users do care if for no other reason than they’re lazy and they don’t (want) to have to create all those relationships and upload their photos – all over again."
Andrew Anker (Six Apart's EVP of Corporate Development and GM of Vox): "Our products (Vox included) serve a lot more than just the web 2.0 crowd and I'm hard pressed to understand why this isn't a global need."
Tim Spalding (Founder of LibraryThing): "Allowing users to import and export their data freely is just another part of providing a good and trustworthy service."
Ben Werdmuller (Technical Director of Elgg and Director of Curverider): "Your web presence is a branch of your identity, and you should absolutely own the data associated with it. In terms of copyright, you already do in most cases - but that means very little if you can't decouple it from the tool that created it."
Read the full interviews after the jump...
Aside from the web 2.0 crowd, do users really care about data export? And if not, should they?
First of all Schmidt is lying, there is no evidence what so ever that his statement is nothing but playing up to the crowd. You’d think by now they’d have access to one’s profile info, provide APIs to move data. Give us one smidgen of evidence, but no, there’s nothing. So just because he said all that, doesn’t mean anything - they have done NOTHING to facilitate this dream.
Users do care if for no other reason than they’re lazy and they don’t (want) to have to create all those relationships and upload their photos - all over again.
Import has been one of the top requested features of Vox, so much so that we decided to make it the core of Ben Trott's demo at Web 2.0 (which ended up syncing nicely with Eric Schmidt's comments). We're going to release the feature to the public in the next week or two. We've certainly gotten a lot of mileage out of the fact that all of our products have always supported full export.
Our products (Vox included) serve a lot more than just the web 2.0 crowd and I'm hard pressed to understand why this isn't a global need. Arguably, the less sophisticated a user is, the more important it is that export not only work, but work easily. Sophisticated users can make multiple copies and keep backups of locked in data. Less sophisticated users aren't as likely to do so.
I am sceptical of absolutism. Allowing users to import and export their data freely is just another part of providing a good and trustworthy service. On LibraryThing, for example, the export feature is a hedge against the company going belly-up. When you've entered 1,000 books, that sort of thing could keep you up at night.
I don't think a more radical, rights-based rhetoric has legs. A few LibraryThing users have asserted a moral right to the money LibraryThing makes off their data, as happens when someone checks out their library on LibraryThing and goes on to buy a book on Amazon. Most users understand that this is a business decision, not a rights one. We enable people use their Amazon Associates IDs for the LibraryThing widget on their blog, but we use our own ID on the site itself. Hey, we built it.
We've found that they overwhelmingly don't. A lot of these services don't store critical data that people desperately need (for example, Myspace or Facebook), and those that do (del.icio.us is one) are well-used and profitable enough that their disappearance isn't a major worry.
I agree with Marshall that it's a rights issue. Your web presence is a branch of your identity, and you should absolutely own the data associated with it. In terms of copyright, you already do in most cases - but that means very little if you can't decouple it from the tool that created it.
All this is subject to market forces. If one or two tools set up true data interoperability with easy-to-use open standards, the others would have to follow suit. This would be a really good thing for the industry: rather than being locked into del.icio.us, say, users would be free to jump ship to a better tool at any time. Services will then need to compete on features not just to get people in - which is often a word of mouth thing - but also to keep them there.
Presuming it's good for users, what's in it for providers? Isn't lock in a proven and successful business strategy?
That’s called Web 1.0. Yahoo has shown that they can make open – a competitive feature. Facebook has also forced the issue.
Proprietary software was a proven and successful business strategy until the open source model demonstrated a number of advantages that changed the game around software development. We believe the same thing is going to happen around "locked in" versus open data. To be specific, the social media industry is based around the idea of so-called user generated content. Why we should make money via locking in a user's self generated data is beyond me. Our users do the work of creating the content and then we don't let them have access to it? We should make money by borrowing a user's data and providing substantial value back to the user in exchange. And as soon as we break that quid pro quo, we should expect them to go elsewhere.
Plus, open data models allow for mash-up type apps of a completely new kind. By opening up data and APIs, users can get creative in taking the data from one site and another and mixing it up for third types of applications that are completely novel. We gain a tremendous amount as providers by letting users play with our app (and their own data) this way.
There are advantages to both, and although openness is gaining, lock-in won't go away. I think you'll see trickery, though. One of LibraryThing's competitors... has an easy export feature. But the legal fine print grants them copyright (not just a license) over any data you put on the site. That's eating the peanut and giving back the shell. Expect more of that.
I can see a time when users become sick of it. As web services become more like applications, more PC-like ineroperability will become necessary: if I want to open a Word document on my laptop, my choices aren't limited to Microsoft Word. I can use OpenOffice, WordPerfect, Google Documents, and so on.
This kind of interoperability will also be a major selling point for businesses. At the moment most web services are selling to consumers, but businesses rightly have more concerns regarding the safety of their data. In the future they may be able to outsource certain applications to a web service provider, but perhaps want to store their mission-critical data in their own datacentre.
Flickr provides an API for data export, which helps users move from Flickr to a competing service more easily. This arguably keeps Flickr 'honest' and provides an incentive for them to keep on improving the service. Is this true, and should it be a proviso that competitors provide the same in return?
Yup - they have single handedly set the standard of excellence.
You bet. I personally won't give my data to a website that won't let me take it back. Why should I when great sites (like Flickr, del.icio.us and Six Apart's own products) don't require it yet still provide great service? We believe users should use our products because we provide a great value, not because they can't leave.
I agree with the argument for enforced honesty. Lock-in can "tempt" a company to self-destructive ends, and ease-of-flight can spur them to excellence.Walled gardens aren't just about competitors. Many LibraryThing users also belong to book-swapping sites. Our users wanted an easy way to link their accounts, but the top sites demanded exclusive deals. LibraryThing ignored them, published the technical directions, and directly urged their users to lobby for integration. We even went so far as to create broken links (broken because, without their data, we couldn't know what books they had) and listed non-partners in reverse Alexa-order to stir up trouble. One by one sites went along, and LibraryThing integration is now a standard feature, supported by eight swap sites. Of course, we passed up the money we would have made on an exclusive deal, but forcing openness made LibraryThing better, and that's worth money too.
I think it is true, and it shouldn't be a proviso that competitors provide the same (although it's good practice that they do). What Flickr has done, however, is provide a proprietary API. Is anyone else allowed to implement the same set of calls on their service?
Standards need to be usable by everyone, and I think the most mileage will come from technologies like Atom or the generalised XML-RPC APIs: if everyone can implement it, you don't need to make special cases for sites like Flickr, and when something new comes along nothing new needs to be programmed. You can program your interface once instead of creating it again and again for every new site that comes along. A proprietary API is its own type of lock-in.
To their credit, Yahoo are aware of this and appear to have made a commitment to standard APIs.
Do you think MySpace and the social software industry as a whole will eventually open up?
I'd guess that there'll always be exceptions but I think already you see a strong trend towards opening up, especially among the newer sites.
I don't know. Social networks have strong monopolistic tendencies. Their value goes up with each additional user, so competitors need to create a lot of non-network value or engineer a sudden "rush" of switchers. I think the war will be lost or won by sites that help people aggregate across networks. Neglecting to provide an export is one thing. Actively blocking an aggregator used by your customers is another. And the more you let them take, the lower your wall becomes.
Most of the social software industry is focused on providing a return for investments that have been made, and as such lock-in and a simple advertising model are probably the most efficient approaches for them right now. However, as the market matures beyond sites like Myspace and Facebook and starts to expand to more useful applications with a social aspect, I think it will become a requirement. It's not going to happen until users demand it, which I think will happen as the demographic becomes more professional.
Thanks to Marc, Andrew, Tim, and Ben for taking the time to give their views on what I think will be an increasingly important topic in 2007.