Chris Saad, chairmain of the DataPortability Project was interviewed by Khris Loux at The Next Web Conference in Amsterdam today. Following is my transcript of it.
Chris Saad: first we have been involved into APML, then we discovered that there are so many standards out there but nobody knows how they might fit together. So DataPortability was born.
What is Data Portability?
Imagine that you can login to several social networks and you have to setup yourself again and again. It might be wishful to move that data around.
Audience likes it and most people have heard about the DataPortability Project.
What is the benefit of all this (for the consumer, publishers and startups?)
Chris Saad: The benefit for vendors is that you get more information about a user than normally. With the consumers permission services can share their information so they can do more useful stuff with it.
Users want to control their data which the silos keep anyway. You should be able to move around your data, control where it goes.
What is the problem then?
The problem is not technical but political. (there was more here but I didn’t get it)
To what extent do you feel the larger companies are sincere? What are the next milestones?
Many big players are moving already, Contacts API etc. are examples. You can also say they only join because of the PR. This is also fine with us but the advantage will be with the new startups who embrace these standards.
Everybody in the room has the ability to out-intervent them and provide better services.
According to Chris the vendors already have done 50% of what they can do for DataPortability. If they don’t go on then others will.
What are the strategic advantages for those who embrace these standards?
Startups: Reduce barriers of entry, you get more data quickly.
Large Vendors: You get a broader picture about your user, you can e.g. import Amazon recommendations etc.
Altogether it is a broader ecosystems and they can work on improving services on top of that.
One advantage for Google et al. is that they have that information alone (competitive advantage). Why should they give that up?
It is similar to private networks in the past. They could choose to not join the new thing „internet“ and back then they lost when they didn’t. The new innovation platform coming up is data.
What are you competing around if you are not competing around your data?
You can compete around user value. You never own users, users own you. Users will have more and more control over their data and where they go.
How can you as startup work on this? What should they do?
Join the DataPortability Project, join one of the Action Groups and join the discussion. Everybody is invited to join. Or join one of the standards groups.
Is it an inside club with all those big names in it like Google, Facebook etc. How do I count as a startup?
The short answer is: No! They have no control. The big ones are equal to the rest. The interest in our group is a grassroots effort.
Which vendor/start is a good example for embracing it?
Microsoft is very engaged, followed by Google, Facebook and digg.com is also quite involved.
The ones who are loosing right now in this area (like MS) are the more interested to join this effort.
What control do you have as a user?
You just hit „No“ to DataPortability if you don’t want this. There will be the default of opt-in and you can control who gets what data in a detailed way.
Who owns the data? Usually if a consumer puts data in your database you own that data. Now you want „I own that data!“. What’s your view on this?
I don’t think there is a mainstream vendor today who disagrees with the idea that you as a user owns the data you created. There is also the fuzzy line with comments etc. which is other peoples data on your data.
There is no point in owning the data until you can control that data. Stuff like revoking permissions, moving it, deleting it, etc.
Khris Loux: You will in the future not having users because of the data lockin because of the service you provide.
Ian Forrester said that if your users are there just because of that lock-in then you already have lost.
Is there a strategy that you or we can implement that encourages Google to move that OpenSocial code into the open source instead of having it in their foundation?
These are different things, OpenSocial and DataPortability. With OpenSocial you can create islands. With DataPortability we try to create bridges between those islands. With DataPortability we have an experiment on complete openness. There are no hidden forums, everybody can be involved and be a leader on some topic in 1-2 weeks.
Audience (Gabe Mac): How can I mix accounts when I am signing up to a new service?
Vendors will do that for you. They can import then data from your existing account. It will all be permission based.
Audience: Are you guys willing to open up your database and implement DataPortability (question to organizers) so that attendees can extract their data out of the system.
Organizer: ConfNetwork is designed for this to interact with each other. But we need to talk about the DataPortability issue.
Khris: There are 3 important aspects for DataPortability:
- Backup and Recovery (if the service goes down)
- Move to competitive service
- you might want to do something more interesting with that data/mashups
Audience: Who owns the data which is the behaviour on some service (like recommendations)
We need a definition of who owns what. These are open questions which are still being discussed. But if it’s data like attention data (derived from your data) then it should be your data. It can be e.g. wrapped up in APML. If you provide proprietary data in the mix or your friends does (like comments). Chris thinks that it might be you because they added it to your stack. But that’s to be discussed.
You are talking about Identity2.0, when is that going to happen?
Identity is data but it is a special kind of data (I am also talking about Data2.0). Identity 2.0 already has already happened because they are already millions of openids out there. The question is why it’s not more widespread on services. DataPortability might make OpenID more relevant and vice versa. It will happen in the next 12-24 months. It only happens if all you guys discuss it, demand OpenID, blog about it.
Audience: This concepts sounds as single-signon across the web. How does the update happen?
You are talking about synchronisation? (yes). Part of the DataPortability stack discussion is XMPP.
Khris: Let’s take a survey here: Would you rather signup to a company which locks your data in or would you sign up to a company which implements Data Portability
Answer is clear I guess :)
There were some different anwers to the opposite question asking about business owners but I didn’t get the numbers form the stream. It might also depend on how you ask the question e.g. if you might want to get data from other services as well or not. Then it might look differently and an open approach might be more useful.
Disclaimer: Not guarantee that I got everything right or this is complete.