The Future of RSS
Posted by Jeff as Uncategorized
May 1st was RSS Awareness Day and while I have a seriously vested interest in this topic, I chose not to observe it. Why? In short, I was busy on other business but also I’ve had this subject gnawing at me and it just took a few days to compose my thoughts.
Let’s take a look at the main areas that RSS companies, like NewsGator, focus on and score the success:
- Blog syndication: This is game over, RSS is the prevailing mechanism by which blogs are syndicated, and a great many people consume blog and traditional media content through dedicated readers or start pages that have generic RSS capabilities.
- Enterprise: Success here has been more elusive. We are doing a robust business in the enterprise but selling a generic RSS platform into large companies is laborious and time intensive. As I have noted before, there is a broad disconnect between IT organizations and how users actually employ technology, RSS is a great example of this divide.
- Widgets: Strong success here, we have been deploying a widgets/data service to media companies that is entirely built on RSS technology. We are currently delivering 200 million widget impressions a month at this point, suggesting that RSS powered widgets are a bright spot.
- Media: Too many media sites put out partial text feeds, this impairs the ability of companies to build next generation consumption applications that rely on some degree of keyword/entity extraction for semantic features.
- Start pages: Great RSS use case here, it is estimated that about 60% of the activity in popular start pages is reading RSS feeds.
- Authenticated Feeds: In a nutshell, this is a disappointing area because the overwhelming majority of applications that could deliver personalized and user/password protected feeds just don’t. These applications include all of the popular enterprise applications and even popular on-demand apps like Salesforce.com.
- Feed Monetization: Mixed results here, FeedBurner enables dropping in of contextualized advertising but it’s not apparent that many people are doing this with great success. The problem with putting ads in feeds is that it’s to easy for client applications to strip out the ad units automatically.
- Desktop Client Apps: Of course we enjoy market dominating success here but it’s not shocking to suggest that desktop RSS apps appeal to a core heavy use audience and growth will be minimal to moderate at best.
- Mobile RSS: There are apps for mobile devices but none have really broken out as success stories. It’s my observation that applying a desktop RSS consumption mode to a mobile device is a nonstarter.
- Attention Streaming: This is something I am intensely interested in. We, as in a small group of leading companies, are just starting to get some momentum here so prognosis is good.
- APML: A promising initiative but while a lot of apps are generating it, very few are actually consuming it. I would also wager that a lot of potential developer partners are probably waiting for APML to move into a traditional standards process before committing. I think this was a failing of OPML as well, lacking a third party to move it forward it just plateaued. A lot of applications use OPML, for sure, but it’s all manual… imagine how much farther along we would be if there was a common feed store that enabled all applications that can consume subscription information to access a subscription, rather than having to update every application with new subscription information manually?
#10 is where I want to start with in terms of continued discussion. Basically the entire RSS market has been built around a use mode of subscribe-then-read, and that is likely to continue as an exclusive model for many users or in parallel to other use modes. The weakness in this approach is that you only know what you know, as in you have know about a feed before you can subscribe to it… and I generally work off the approach that it’s far more likely that the best content on any keyword is not necessarily found in my OPML.
There are an increasing array of companies that are working on a next generation of feed consumption use model, built not around the explicit subscribing of feeds and chronological consumption of content. In order for RSS to get to the next level of mainstreaming we have to think in terms of behavioral filtering of content and discovery of new content sources based on explicit preferences or inferred preferences derived from behaviors. This is exciting for me as a user.
I think one of the reasons why Techmeme has proven to be a consistent favorite is that this next generation model is partly how Gabe built the system. Through using Techmeme I am essentially outsourcing feed discovery to the service and consuming content not based on subscriptions but topics. As a users, ordinary or power, I would like to have a personal Techmeme that delivers content based on my consumption habits, or put another way, my attention streams.
To further develop this model, I would like to see a social dimension develop that pushes up/down content based on a collaborative filter that takes into account my social graph and what they are consuming and rating, explicitly or otherwise. The problem with rating that we need to overcome is that a very small percentage of people will actually score content, so that’s why the attention streams become valuable, through activities they are effectively scoring content.
For enterprises this is nothing but goodness, but unfortunately it will likely be that enterprise users are the last to benefit from these advances because they are dependent upon IT. It will happen but the use cases we have to build to will be specific and in some cases tedious in an effort to get a flywheel spinning that elevates RSS in the enterprise to a strategic focus.
Lastly, RSS is an ideal syndication technology for publishers to take advantage of to expand audience and monetization of content. The missing link up until recently is widget technology, because inserting ads directly into feeds is a nonstarter in my opinion. Charlene Li said it best when she said “wigets are RSS in a dress”. Charlene, who attributed the quote to someone else, wasn’t being facetious or otherwise inappropriate, she was observing something deceptively simple, widgets present a use mode that average users like because it puts them in control while presenting a visually compelling experience.
With traffic acquisition costs being what they are, media sites are unlikely to see significant growth in site traffic using linear techniques like adding more pages and/or better SEO. “Going viral” is popular to say but hard to do, widgets offer a simple and low cost mechanism for publishers to use to drive additional site traffic not by getting site traffic but by dispersing content to a variety of destinations, including social networking pages, blogs, start pages, and non-affiliated websites.
On balance RSS has been an exciting area to mine away in, with a diverse range of target markets, use cases, and a broad array of interesting companies. It’s here to stay but at the same time is begging for a new interaction model to drive broad mass market adoption; it’s still too geek. RSS in media is the ideal syndication technology and with newer technologies like widgets, finally able to be monetized.