Locked-Up Data for Web 2.0?

I have not poked at a blog posting of Matt Asay's in quite some time, so no time like the present. This morning I ended up reading a post he wrote for CNET on August 20, "GPL is the new BSD in Web2.0, and why this matters." As always Matt is pulling at the threads that tie together the big picture elements - him being extremely smart and whatnot. The target of his comments was really about source code licensing and how that applies to the world of Web 2.0 (and beyond).

I agree that the web is driving outrageously cool technology resulting in new business models all the time. And, I agree that these changes are going to adjust the way we think about what "unique value" really means in an environment dominated by the network effect. And, I agree that the legal constructs under which we have been functioning for so long are going to be challenged and stretched. And, most importantly, I agree that in the world of Web 2.0 data is king thus making the network effect so powerful. Ok - I got all the agreeing out of the way with that.

Here is where we fundamentally part ways. The reason people don't get over the "software fetish" is because that is still a huge asset - and it tends to be the really expensive part of creating any services offering. Data is the other massive asset - but you don't get that without high value software underneath. People will continue to value their hard work in producing that software.

Open source projects with bad code...are still bad. Proprietary projects with bad code...well, you get the picture. I have had this conversation with Matt in the past, and we will continue to see this issue differently. Fine.

The thing that really surprised me in the posting though was these two simple sentences, "Lock us in through data. Fine." 


I have been talking with executives, government officials, academics, etc. all over the world for the past 2 years about data. If there is one thing that people REALLY do not want locked up by vendors it is their data. Online, enterprise apps, consumer devices...nope - don't lock up my data. In fact, this concept has catapulted the rather arcane world of data formats to the top of the industry news heap lately.

If I had to choose between buying a service online from a single vendor who believes their value is in their software, but get to control my data vs. using an open technology where my data is locked up unless I pay for it...I think I want my data thanks.

As I sit and write this my own hypocrisy comes filtering into my thoughts. I have been paying a subscription service for music lately - and don't own a song. Hmmm...so there is a place where I am tolerant of someone else controlling my data. But, I don't think I would feel the same way if I was creating the music vs. just consuming songs created by others.

I don't think we are going to wake up in 20 years and think about how goofy we were about software. I think in 20 years the development tools, and environments in which apps (or their future analogs) will run are still going to prove that he who is able to build the best code (innovation) that delivers a great service (business model), is going to be making some good money.

What I think we will find in 20 years is how much more people value their data. So much more of our lives will become digitized and that will lead to greater awareness of what data is ours, and how important it is. Just think about the progression of the privacy issue over the past 15 years. People want to control their data, they expect the companies building solutions that they use to enable that control while still offering powerful, compelling technology qua software.


I deleted a paragraph that I had been drafting and decided not to pursue - I goofed up by having it still on this post...


Comments (6)

  1. Matt Asay says:

    Jason, good points all.  I only fault you for not reading far enough.  Had you kept reading (the blog, not that particular entry), you would have stumbled across this post <http://blogs.cnet.com/8301-13505_1-9770425-16.html&gt; (the importance of keeping data open) and this one <http://asay.blogspot.com/2006/05/future-of-lock-in.html&gt; (Sharepoint as the center of data lock-in).  I 100% believe that data should be open.

    I also, incidentally, believe that software should be open, and think that this is especially critical in the Web 2.0 world.  What good is my data if I don’t have the software that it runs on?  Let’s say I leave Salesforce.com as a customer tomorrow (I’m not a customer, but follow the reasoning)…what am I going to do with that data, even if I can take it?  Not much.  But if I’m with SugarCRM (I *am* a customer), I have the source code (most of it, anyway) and the data.

    SugarCRM is clearly a better option from the anti-lock-in angle.

    So, anyway, I think we actually agree on this.  You were just disagreeing with an early post of mine when I was arguing a different point entirely.  But sophistry, thy name is…. 🙂

  2. Chris Clark says:

    Jason, just to comment on a couple of points here…

    Open source projects with bad code…are still bad. Proprietary projects with bad code…  Could I just highlight that the target of choice for hackers and spam emailers is always Microsoft software – 298 spam mails today and counting.

    I am very pleased to see your views on avoiding data lock in – it was only last year my Microsoft Money application sent me a message saying upgrade or expire, so I didn’t upgrade, and it expired taking my data with it.

    Last point is Google uses pretty much universal open source software – now that is bad open source clearly.


  3. jasonmatusow says:

    Matt – part of what surprised me so much was the fact that I had not seen you take that angle before (in fact, as you say, quite the opposite). I think that there are many sides to the example you make of Salesforce and Sugar. Each company has to make decisions about where the value of their offering is, and how they both use and protect those assets. That is the vendor side of things. From the consumer it is a question of migration pathways in terms of the long-term access to the data. In the short-term, users may well choose other elements to value in their use of a solution.

    Having the source code – marginal value at best for the overwhelming majority of people who ever use Sugar or any other solution. If I can export my data and drop it into another solution – you will have a hard time convincing me that it is more efficent (from a monetary perspective) to stand up, sustain, and modify (to meet future needs) my own implementation of Sugar than to take the short-term costs of migration to a new solution supported by a different firm.

    I think we agree on the data thing. I don’t think we do on the source code-openness-as-value-for-customers thing.


  4. jasonmatusow says:

    Chris – thx for the comment – good stuff.

    I would point out some things though in response. First of all, IMHO you are mistaken in your conclusion about spammers. Spammers are attracted to the broadest possible audience as their target. It has nothing to do with the technology. This is true for hackers too. The security game is almost entirely about the quality of the target – not the underlying technology. All of the metrics I have ever seen about attack patterns support this. The resulting vulnerabilities could well be argued pivots on the technology (we’ll ignore how the IT staff deploys, puts in place policies, actually enables security features, responds to fixes being made available…).

    As for Microsoft Money, I am sure there is an export feature from the database. My guess is there was a whole bunch of export options. Also, I would assume that other personal accounting packages have import features that would accept that data in a number of ways as well. Worth doing a search on TechNet about that one if you have data you can’t get to.

    No, Google does great work. I have no idea what your point is with the last comment. I hate to burst a bubble, but there are bad open source projects – just like there are bad software development projects in all models. My real point was about the value of software and what it takes in terms of investment to make good software no matter how it is developed.



  5. Jason – I have to agree with you on this.  Companies want their data, and they HATE data lock in.  The key for most companies is just what you describe, the ability take their data with them and use it in another application – not having the source of the software.  I have seldom seen a company that would have a clue what to do with the source code, but they all know what to do with their data. – Ben

Skip to main content