I remember when I first heard about Twitter, which was early spring 2007. There was this enormous buzz about this new cool site where you could update your friends about what you were doing and when â€“ all this via text messaging. At first I did not notice the site at all as text messaging is old news in Europe so it didnâ€™t seem very new. At first it actually sounded a bit obnoxious and it really was at first. However it matured and here we are in 2009 with a highly popular site used for a lot of different purposes.
First of all and importantly, Twitter has never really positioned them as cellphone powered news. We every day hear about how Twitter is growing in popularity and I myself am a big tweeter so I am not out to bash Twitter. It is a really cool platform for quick little link exchanges, updates and discussions with your friends. Others have however claimed that Twitter is a news site.
Twitter is not a cellphone-powered news site. It is a tremendous source for what could become news, and a resource for journalists, but on itâ€™s own, it is not.
There are two main issues with Twitter from a news perspective:
- It is hard to find things on Twitter.
- The tweets are not validated, or fact checked
- It is very much a one-perspective story, and the short format is not favorable here.
Let us first take a look at the first point here. The intention of Twitter is not to make things searchable. It is a cellphone powered social network to stay in contact with your friends and others. There will be other applications as natural extensions but primarily Twitter was built for friendship discussions. As all platforms people play with them, but the fundamental architecture is built for interaction only with your friends.
But hey now, Erik! There are search engines that let you search Twitter. Sure, but they are not on Twitter. I get very little support to find other content that is outside my social sphere. It makes the discovery part of news consumption pretty much impossible. Your consumption will be as strong as your community.
I have to admit here that the news consumption is changing and the media arena is moving more towards gatekeepers who help us stay in touch with our news. Before the media newsroom was that, but now the consumption is relocated to the spheres around the information hubs in the social network. The social media reaches the journalistic arena. This is the right development and pretty sweet of course. You still need to assist the users with other tools to navigate through the content and find content. Nope, hashtags will not solve this. :-) They are bandages, not a solution.
The second and third issues are related. One of the main challenges with user-generated content is to validate and guide your consumers whether a specific piece of content is credible or not. The need cannot be underestimated and I have seen different approaches here, but clearly the content creator will not be able to solve this for you sufficiently. The content creator should be free to publish and we as a news industry should do assist in the guidance of the consumer of whether a particular piece of content is credible or not. The length of the message is not playing in favor here as 140 characters
Can Twitter overcome this? They might. They might not. Ultimately I do not think this is in their interests, as it would kill the fundamental idea of Twitter.
Ok. So what is the solution then?
The media scene is continuously moving more towards the ultimate content mashup on a topical basis. The articles are no longer essay like, but more a mosaic of different media types, angles, presentations and such. Essentially, you will have three main components to the presentation:
- The Report – The report is the foundation of the media experience. Anyone can report from anywhere via cellphone or PC, sharing news, images, videos and opinions.
- The Related Voices, Opinions and Facts – Around the report multiple perspectives are woven together that is have a collection: news stories, blog posts, images and videos via contributors and aggregation.
- Discuss and Rate – For each of the presentations you will enable the community to emotionally connect with others around the world through discussion, rate the story and in the community spirit complete the human story.
Key in all of these components is the freedom of contribution, interaction and discussion. It is an absolutely necessity to have this free, and really a natural consequence of freedom of speech. It has to be free and open. To all people. By any means. Everywhere. Essentially, all these components build the human story and the provided context, facts and opinions built up by other user contributions and aggregation grounds the discussion. Ironically all these characteristics of the content actually provide you with the necessary components of the quality assurance system. It is the ultimate long-tail system.
First of all let us kill a myth. The long tail content market, especially including the user generated content, has always been the subject of a broad and intense discussion about quality assurance, especially since most of the solutions proposed includes heavy emphasis on technology. Traditionally human editors have been used as the major and gate keeper part of the content quality assurance process. The use of hired human editors is not a scalable solution and never will be a scalable solution. In addition the traditional editing model via human editors has also been proven to have some concerning issues especially for high profile news stories/events. The vast number of surfaced staged photographs published by for instance AP, Reuters and New York Times via their photographers during the conflict in Lebanon 2006, as well as the plagiarized and fabricated stories by the New York Times-journalist Jayson Blair have really seeded doubts that human editors/journalists are really the solely solution on the long term. This problem is also seen in the new social media sites and one of the more important examples are the issues with the super editors in Wikipedia.
The concept of content ranking â€“ yes even PageRank – have always struggled between diversity and singularity. Usually they turn to the singular answer. This-is-how-things-are solutions have ruled the content ranking scene. Maybe the paradox is us as humans. Singularity is simple. It is transparent. Diversity is not. It creates this twilight zone reality where you do not really know what is the true or false. The easy, simple, neat answers disappear. PageRankâ€¦ The amazing and brilliant algorithm that disrupted a whole business and market. It gives us what we need. Or does it? Really? PageRank works (simplified of course) very much like biological evolution. The strongest (or here most reputable) survives. The strongest win. They conquer the weakest. The alternative voices such as the extinction threatened species are not heard. They are lost in the noise. Is this right? Or is it â€œjustâ€ the natural choice? Hard question right. It ties into the news worthiness of a piece of content. PageRank type os ranking solutions are designed to give you one answer, but do not adequately rank in the long tail market where you would like to present clusters of information. At least not without modifications and conceptual changes to the overall nature.
Clearly we are looking at a solution that is simple in itâ€™s core – technology heavy yet community and participation intense. It turns out that the main features of the solution should be:
- The reputation of the contributor
- The collective power of the aggregation around the user generated content
- The community interaction with the content such as viewing, rating, commenting and other contributions
It is obvious that the right, sustainable solution will be a combination of these three ranking attributes. Basically you turn your challenges into assets in the editing process, as well as use them as opportunities to address the business concerns. This combination will be much more scalable, much more robust, completely technology based and still addresses all the issues with the ranking of user generated content. This is especially true for the long tail market where the amount of content is just too big for humans too handle – both practical and consistently. Think of it as a simulation of the traditional newsroom process yet with less bias. Isnâ€™t that cool?
Why will this work? Easy. :-)
A contributor sends in a text message, which gets instantly published. Now by extracting keywords from the message related content online â€“ news stories, blogs, images and videos â€“ can be associated with the text message. As the context around the message builds up the community interaction with the content is constantly growing.
During the whole life cycle of the message we can define the credibility of the message at any given time and guide the user. The important conclusion here is that we are not reliant on a single point of failure here. If the contributor is unknown the other two components can take over. If we do not find any related content online we will be able to use the reputation and the community for validation. If the interaction from the community is small (which for the long tail it might be for some content) we will be able to use the reputation of the user and the related content to assess the credibility.
Oh wait a minute. That will help us tremendously to strengthen weak voices, detect breaking news and also complete and broaden the story telling of a particular piece. I am pretty stoked it already exists.
Always remember that the future media scene is like a dinner table conversation. It is free, open, interactive and human. I usually describe with thinking of a party. Which party would you like to go to: the party with all the detailed instructions what to wear, what to do, what to not say, what to say. Or would you prefer the laid back come as you are party. I would choose the rock n’ roll party any day. Why? It is fun. It is free. Butâ€¦ Most importantly I can relate to it. I can relate to it. That is the key. Sure, you will need some logistics for the rock n’ roll party too, but the script is not already set.