Living On The Edge (the rise of the edge feeders)

I believe that the value in any network is created on the edge. 

That's where my blog sits.  That's where Gotham Gal's blog sits.  That's where Ted's blog sits.

That is not where Google Base sits.  Google Base is in the center.

In the freebase model, you give your content to Google and they keep it.

In the myspace model, you create your content on your page and you keep it.

I am for the myspace model and against the freebase model.  And so is most everyone I know who has spent anytime thinking about this stuff.

Michael Arrington says this on CrunchNotes this morning:

Google Base has a lot of problems. The big one for me is that is another centralized content play, when all of the really interesting content is created at the edge. Jeff points out some other issues with Base, including the fact that the data is not only centralized but there is no API for non-Google search engines to access the data.

Jeff Jarvis says this on Buzzmachine this morning:

If I were a VC, I’d be investing in a company that tries to use tags and microformats and social interaction to link together the topics and opinions and information people care about on that distributed web. For that’s the company that won’t waste effort and expense trying to get people to change their behavior and reverse the natural flow of the web out to the edges — ‘come to us and give us your good stuff’ — but instead takes advantage of the essence of the web and leaves control out at those edges by saying: ‘We know you have good stuff and we’re going to help people find it.’ The consumer proposition is then clear: This is how you find the good stuff. This will be the real successor to and competitor against Google. Oh, Google could do it, too, but judging by Base, they’re not doing that. They’re taking control rather than giving it.

Well I am a VC and I sure would like to find "the real successor to Google" so I pay attention when someone as smart as Jeff is giving free advice to me and my colleagues.

But as a content creator living on the edge, I am not sure we are ready for microformats and tags and social interaction to do all the work for us.  We'll get there, I am sure of that, but we need an interim step and that step are services that feel centralized but are really application specific edge feeders.  I need a better word for these services, but for now I'll call them the edge feeders.

What edge feeders do is provide mechanisms to faciliate the content creation on the edge.

Flickr is an edge feeder and the best one I know of.  I could take my photos and simply post them to my blog.  But I don't do that.  I put them on Flickr and then from Flickr, I post them to my blog.  Flickr makes that dead simple.  But they also give me a badge to show aggregated photos. And they let me post other's photos to my blog.  They are the photo feeder of the blog world.

Blip.tv, vimeo, and youtube are the video feeders.  I could post a video directly to my blog, but I don't.  I post it to vimeo, youtube, or blip.tv, and then from there I post it to my blog.  These services are rolling out lots of video specific blog integration techniques that will make it even easier to be a video content creator living on the edge.

Delicious is a link feeder.  I could post a linkroll to my blog, but I don't.  I use delicious to host all of my links, and I use a tag (mine is linkroll) to feed my linkroll.  Delicious makes it easier to be a link content creator living on the edge.

And the service that got Jeff, Michael, and me talking about all of this today, Riffs, can be a feeder too. Jeff points out in the same post that I quoted from above that, "Gotham Gal has all kinds of reviews already on her blog."

Yes, Gotham Gal writes a lot of reviews.  But its a pain to do that.  She's got to find a permalink of the movie, restaurant, record, book, whatever.  She's got to find a thumbnail.  And then once she's done all that, she can blog it.

A review/opinion feeder will do all that for you. When I review a record, I go to Amazon because its the one place I know that I can find a permalink and a thumbnail.  I use Amazon as my review/opinion feeder.  But its not built to do that. It has no blog integration other than what Typepad has built for them on top of Amazon's API.

Riffs has some work to do if it wants to be a review/opinion feeder but that's an opportunity that is worth pursuing.

The other thing these feeders do, as pointed out by Bruce Spector (one of the people who built Riffs) in his comment on Jeff's post, is that feeders also make great lightweight blog platforms for those who may not be willing to make the committment to seriously blog.

Mark Ghuniem has an amazing blog that is built on delicious

Missinfo has an amazing video/audio blog that is built on blip.tv

Thomas Hawk has an amazing photo blog that is built on Flickr

Some of these people have blogs, some don't, but these application specific blogs they've built on these edge feeders are really compelling, more compelling than most blogs I come across.

So to summarize, centralized is bad, really bad.  It won't work in the world we are headed into where all the value is created on the edge. But just because a web service has a centralized organizing structure doesn't mean it is centralized.  The key is what they do to feed the edge. If they can become an application specific edge feeder, and the best of its kind, then they will help facilitate a massive amount of content creation on the edge and will also build a centralized organizing system (like Flickr's interestingness) that can create enormous value.  If the centralized organizing system itself can be distributed out to the edge, then its a home run.

In reading this over a couple times, I hope this isn't too dense and obtuse.  It makes a lot of sense to me.  I hope it makes sense to you too.

Comments

[Al]

Ok Fred I'm going to take the bait on this one, perhaps go out on a limb a little.

I think your thinking about Google Base (GB) is a little Web 1.0. I think the database is actually designed as a metabase i.e. a database about data rather than the data itself. In that sense google is not becoming a content provider. Rather it is providing a mechanism for use created metadata. This metadata could be something physical like a classified car sale in which case it could only ever be metadata as the car is physical, the classified is metadata around the car and its offering/sale. In fact googlee search features are just metadata, when you search google you are searching their inverted tree word/page map,and it is then passing it through weighting algorythmns based on their ranking metadata. It is then presented with their advertisers metadata and it goes on adn on etc.. Web 2.0 is about metadata, especially socially created metadata.

Now if your saying they are locking this metadata away from competitors, well yes, just as they did with search, just as everyone did with search. Although actually they might just allow you to search this metadata this time. You can use the GB search to do this and I strongly believe that there will be an API following (Maybe atom based). Check out recent posts on Folknology about where we see GB going.

I would also suggest that Flickr, delicious et all are no more edge than GB on this one, they lock away (at least in part) their metadata, only allowing access in fragments. I would read what some others are saying about GB, I think it's actually a very clever play, also check out Adam Bosworths talk from the last MySQL conference for clues about it's origin and possible purpose..

regards
Al

[lawrence coburn]

I think that makes a lot of sense, though I would argue that some centralization isn't all bad.

For example, your edge is my centralized destination site.

Specifically, anyone who comments on your blog posts is complying with a centralized model where your blog is the center.

So why would we do that? Because there is some value, I think. There's the conversational aspect (mentioned by Scott Moody elsewhere), and there's the audience that you have built over the years - two things that are very important to me as a content contributor.

I don't think it will ever make sense in this sort of setting for me to post my reactions to your posts on my own site (MY edge), as opposed to here. It's just not conversational to do it that way. Plus, readers of my blog aren't particularly interested in this sort of stuff, while yours are.

But if your blog would become an edge feeder where each of us could easily feed our contributed comments back to our own sites... ah yes, I guess that's the whole point of your post.

So getting back to Riffs and sites like it, like you, I wouldn't be so quick to dismiss it solely because it is centralized. Centralization and community go hand in hand...the key is to be able to provide the tools to let your site's content contributors take their stuff with them.

[steve]

maybe i'm way too simple, but how about if we avoid loaded terms like "centralized" (which suggests scary big business) and instead use more descriptive accurate words like "aggregating"?

all web sites are aggregators, i'd argue.
and everyone is hand-wringing over Google Base because Google's power as an aggregator has everyone seeing their precious small startups looking irrelevent.

(used to be we all invested our sweat and capital in doing the big guys R&D for them then selling it to the big boys for big bucks when they realize its value. Google does their own R&D as smart and fast and visionary as a startup.)

eBay is the ultimate aggregator. ain't no way their business model would benefit if they somehow opened up their API or offered tools for tagging or invited in screen scrapers or whatever.

ditto amazon and all ecommerce sites

ditto nytimes and publishers

all web sites aggregate something -- information, physical objects, data base entries, people -- then seek to charge somebody -- visitors, consumers, advertisers, etc. -- for access. i cant think of any exceptions

so in the end, i find it hard to believe that the "aggregator" business model (whether you aggregate blogs, photos, videos, data, anything) is going away, or even can be bettered -- and that model is "gatekeeper". and for that you need "gates" and "keepers". so all this chatter is really just everyone going around and around and around on what gates and fees are cool or fair or reasonable. zero gates may feel cool and egalitarian but it requires us to abandon the foundation of property rights which underpin our basic tenets about capitalism and democracy. call me names -- its happened here before ;) -- but for that reason, i don't see it happening

[David Gibbons]

I'm with Al; in this context, I don't see the difference between del.icio.us and Gbase - they're both central depositories of data - about data.

The main difference I see is that Gbase relies less on the audience and more on the author to index the web - but in effect it promises the same result (i.e. your site, categorized) - just with a fraction of the work that it takes to have the audience do the categorization. In addition, Gbase lets users create content - that's just smart.

In today's automated internet, I must admit that despite the "control" that comes with tagging, I was surprised to see "manual indexing" of the internet become so popular. Is this a fluke?

Ahhh - but what about "life archival" and "social algorithms" you say? Well, no-one said those should be manual processes either - do we really need to manually tell a system when we've read a web-page? Do we need to manually click through to other users who read the same page to personalize our search results?

For publishers, tagging is great, it lets me create lists of similar stuff in my side-bar - but will it ever compete with goog? - no.

I'm all for Peer Production, but there should be a nett increase in productivity or participants will find something else to do with their time.

Even at the edge, it's nice to have some things done for you. The internet's getting too big for me to tag it all by myself - I just wish I'd beaten G-base to it - simple - brilliant - and obviously disruptive!

[Thomas Hawk]

Hey Fred. Thanks for the mention of my flickrstream.

The thing I really like about Flickr is that there is already an audience there of people who are community minded and passionate about digital photography.

While I post my photos on my regular thomashawk.com blog as well as flickr (and actually get more traffic at my own blog because for some quirky reason Google Images ranks my photos very very high), it is the community aspect of sharing on flickr that is of the most value to me. To be able to get the instantaneous feedback on Flickr from other like-minded photographers is very precious fuel and serves as great inspiration.

At it's core, it is the community aspect of Flickr that makes it a superior blog for photography for me over my own personal blog. While you give up some control issues (style format, html, custom features, the ability to track referals, etc.) what you get in return is an attractive sleek simplified space (I think of flickr's design as almost a virtual art gallery or museum) to host your thoughts and images at a very inexpensive price for the bandwidth that you might consume. You also get an instant audience for your work and an internal rank system (interestingness) that brings new viewers to your top images regularly.

Flickr in turn gets an annual subscription fee plus ad revenue for ads shown to non-pro users.

In my mind this is a win-win-win. I get something great, people get to see my work, and Flickr makes some money.

Further, I've made many real in person personal friendships with the people that I've met through Flickr -- friendships based on a shared passion that has been very deep. A worthwhile experience that I'd recommend to amateur photographers everywhere.

Thanks again for the link.

Tom

[Terry Steichen]

Fred,

I've re-read your "Living on the Edge" post for the 10th time. I've finally got a handle on my thoughts on how an edge-based company can succeed, and would appreciate your reactions.

1. the company provides a platform and promotes it.
2. edge creators use the platform to produce content.
3. the company imposes minimal controls on content contributors, stimulating (2)
4. the platform enhances the creators' individual content, creating value for the creator, stimulating (2).
5. the platform also produces aggregate content, facilitating the network effect, creating more value to the creator, further stimulating (2)
6. the platform makes some/most of the content freely available via API's which has several effects:
a. expands the reach of the offering
b. defuses potential competition (why build it if it's available for free)
c. defuses potential opposition to revenue generation by company (via ads, etc)
7. the company starts serving ads, and/or provides premium (paid) services

I think the above model describes Flickr (content=photos), delicious (content=references), MyTube (content=video), among others.

Another point is whether a blog is a centralized or distributed content publisher. The way that I see it, a blog (such as yours) is a kind of hybrid. Most of the value comes from the postings (by the blog owner or invited guests), augmented by the value of the comments (from the edge). Thus the blog organizes the content centrally. It also publishes it centrally, for the most part. Readers either go directly to the blog's site, or get the content from the site indirectly via a RSS reader. Some of the content gets pushed out (via RSS) to be republished at other blog and/or newspaper sites, but these are really central providers, not edge.

[Ofer Luft]

Hi Fred.
Xlnt post.
Wanted to show you a more "structured" photo blog, which I built on Flickr.
I documented my trip to Singapore, by using a simple convention of manually entering the date and time of when the picture was taken and uploaded them in chronological order
(geocoding was too much of a hassle to do manually).
Too bad Flickr doesn't allow advanced sorting options.
It starts here -
http://www.flickr.com/photos/oferico/56335023/
and can of course be viewed in reverse here
http://www.flickr.com/photos/oferico, where it's so much nicer to see all pictures and texts on one page.

Take care,
Ofer

[Graham]

Just recieved an email from Google, my product feed was not in the new Google format. Try again. .