IAC and the social web

Reading Time: 6 minutes

This post was inspired by the recent announcement that IAC  will be closing down Bloglines on October 1, 2010. The real question was ‘when’ rather than ‘if’; as the product had been left to languish and maintenance of it had been a low priority for a while at IAC’s Ask division, something that Ask themselves alluded to when they talked about Bloglines “loyal and supportive (not to mention patient) user base“. The post threw up some other interesting elements in there however.  But before we delve into those points, it seemed right to give Bloglines a fitting send-off.

Requiem for Bloglines

Bloglines was set up over seven years ago. I liked it for its ability to handle a large amount of feeds through a three-window view which would be similar to anyone who has used an email programme. Having the RSS reader on the web rather than on a computer meant that I could catch up with the content wherever I was and didn’t have to worry about clearing out read content on work and personal computers.

Bloglines has a simple interface which works and is more productive to use than Google Reader, for the past three years or so it has had a decent mobile web interface as well. As a product Bloglines was always a bit of a curates egg in the Ask search portfolio – Bloglines attracted a web-savvy or even early adopter user base which contrast sharply with Ask’s senior soccer moms and empty-nester audience base.

I have used Ask for over five years, I originally used NetNewsWire RSS client on my Mac and wanted something online that I could use at home, at work and on the move. I am currently looking around for an elegant replacement, I find Google Reader poorly thought through in terms of providing a productive user experience and a decent experience on a mobile device.

My current thoughts on my best migration option is probably to use NetNewsWire as an interface for a Google Reader account, though BlogBridge would also meet many of my needs. Alternative suggestions would be welcome however.

Ask’s future direction

There were a couple of paragraphs in the blog post which spoke specifically to Ask’s future direction:

Primarily, it’s about focus. Our focus here at Ask is on building out our Q&A offering. Our beta, released in a very limited fashion at the end of July, is seeing steady community growth with an answer to question ratio of 2:1, which underscores the fact people are motivated to provide answers. Another encouraging metric? On average, users who are active within the community are asking 2 questions a day. These early numbers build on the competitive advantage we already have in the marketplace – the fact that 30% of the searches on our site are already in the form of a question – and are further proof Ask needs to continue its tight focus on delivering the best possible answers to users.

A little perspective: when we originally acquired Bloglines in 2005, RSS was in its infancy. The concept of “push” versus “search” around information consumption had become very real, and we were bullish about the opportunity Bloglines presented for our users. 
Flash forward to 2010. The Internet has undergone a major evolution. The real-time information RSS was so astute at delivering (primarily, blog feeds) is now gained through conversations, and consuming this information has become a social experience. As Steve Gillmor pointed out in TechCrunch last year , being locked in an RSS reader makes less and less sense to people as Twitter and Facebook dominate real-time information flow. Today RSS is the enabling technology – the infrastructure, the delivery system. RSS is a means to an end, not a consumer experience in and of itself. As a result, RSS aggregator usage has slowed significantly, and Bloglines isn’t the only service to feel the impact.. The writing is on the wall.

I found this fascinating, the focus on a Q+A service is in some ways Ask stepping back to its original proposition where the audience was encouraged to do natural language searches as questions which Ask would try and answer. However this time it is relying much more on people than algorithms. There were some key points that struck me reading these paragraphs and wanted to expand on it further in three parts:

  • The nature of RSS in social conversations
  • The real challenges that Ask will face on its new Q+A offering
  • The paradigm shift happening around knowledge search and the likely impact it will have on products such as Ask’s Q+A service
RSS is still relevant

Ask are right to point out that the serendipity of social is a way that consumers are discovering interesting content. But there will always be some people who want to follow content from particular sources and some stuff recommended by friends.

Herein lies the challenge, although social is a great way of discovery RSS is still the best way to feed regular content consumption. In fact, that is why the Guardian newspaper publishes the full article in its feeds. I still get the bulk of my blog readership from RSS aggregators rather than through the web.

RSS won’t work for disengaged content audiences, Twitter and Facebook allows them to pick up what ever content happens to float by. However RSS is immensely important for those of us who put those links into circulation. It is also a measure of the failing of conversations to provide a reliable diet of content that Wadds and I were recently talking about the potential role of media as a content curator last week.

I think that conversations have sparked more interaction online that adverts can vended against with varying degrees of success, but RSS still has a lot of life left in it.

Q+A is hard

Ask is right to be happy about their new Q+A service, however these results are coming from a very limited beta testing process which doesn’t really give you a feel for what full implementation will be like. Chances are that the beta testers are likely to be more highly engaged due to the selection process that Ask had uses, what often works in theory falls down in practice. One of the things that I learned at Yahoo! is that knowledge search projects like Ask’s Q+A are hard. What do I mean by hard? They are certainly easier to scale from an infrastructure point of view than say search, but you have to design that scale in from the start. The challenges are three-fold: culture, community and monetisation.

Culture is of a critical importance because it will affect the kind of answers that you get and the relative amount of answers you get. The gold standard of knowledge search services are Naver in Korea and Baidu’s relatively new service in China: both of these exist with a society were harmony is prized, this means that you get more better quality answers. Lycos’ now defunct iQ service which was knowledge search on steroids combining it with social bookmarking. A powerful product that needed some user experience issues worked out. iQ was originally launched in Germany were it managed to garner high-quality answers, however whilst the quality was high the average number was low as Germans wouldn’t attempt an answer they didn’t definitively know.

The old adage of only being as good as the company that you keep also applies to the social web. One of the problems that Yahoo! had with the quality of the responses in Yahoo! Answers was the kind of people who responded. They were generally time-rich and cash poor. This included receptionists, security guards, college students, housewives and schoolchildren. This wasn’t flickr where the team had carefully nurtured a community that continues to carry the image social network along, this mean’t that you ended up with a lot of silly, useless or downright offensive answers. If you have the wrong type of community you have to do a lot of expensive moderation work.

Whilst Ask’s older soccer mom demographic may certainly be people that value having their questions answered I would be concerned about its ability to get good people to give them the required content.

The quality of the community also has a bearing on advertising prices as well, which brings me on to the final challenge of monetisation. When I was originally told about knowledge search in an internal meeting at Yahoo! I thought that the service could disrupt the business models of people like Rightnow Technologies and Transversal who both provide automated product | service support for companies; instead Yahoo! could offer companies sponsorship of particular channels.

Things didn’t work out quite that way and Yahoo! struggled to get the full benefit of monetisation from Answers. Google’s Answers product operated on a slightly different model where the person asking the question paid for it to be answered by researchers with Google taking a 25 per cent commission as broker. Google eventually closed this down due to limited demand.

Paradigm Shift

Back in the day search was something that was bought off companies like Inktomi by media and portal companies. Its big black appliances with the tri-colour nested cube logo was a common site in large media company datacentres and ‘internet hotels’ of the time. Overture came up with the model of search advertising and Google realised that consumers were willing to move for a great search experience and nothing but the search experience. A radically different way of doing things can shake up a whole industry on the web. I think that the same thing is happening with knowledge search at the moment. Hunch is in the vanguard of this move, interestingly it has Caterina Fake at the helm who realises the importance of culture, community and monetisation from her time at Ludicorp | flickr and eventually Yahoo!

Hunch changes things as it focuses on using decision trees and providing users with answers by tapping into the site audience’s creative intelligence. It has a ‘game with a purpose’ feel to it. Whilst it isn’t the next Facebook I think it will change the way we think about knowledge search because Hunch allows the service to really dig into user intent.