Ethan Zuckerman wrote a very interesting post today called What if Search Drove Newspapers? He talks about several different initiatives designed to gauge readers’ interest in different news stories, particularly those that are currently under-reported, and then devising methods for reporting stories on these topics. He asserts (correctly, I think) that this is basically search-driven content development. In particular, this is a strategy that will work well with Google.

Zuckerman concludes by making an interesting point (but you should go read the full post):

I’d propose another way in which search-driven content creation might be evil – it’s a step towards news outlet as search engine and away from news outlet as source of serendipity.

The front page of a newspaper is a statement not just about what’s happened in the world in the previous 24 hours, but what the editor believes is important for you to know about. There’s always more that happens in the world that can fit on a paper page – or even a much larger web page – and the editorial decisions made shape a vision of what you need to know as a reader and what you can safely ignore. Smart editors use this ability to engineer serendipity, pushing readers towards topics they might not have known they were interested in, featuring more obscure content that’s got good storytelling and a high likelihood of capturing a (previously uninterested) reader’s interest. (I wrote about this idea at more length in a post called The Architecture of Serendipity.)

The way to create value in digital business models is by creating value through aggregating, filtering and connecting ideas. The thing that I think is interesting about Zuckerman’s piece is that it basically looks at Google-style filtering as the only method for driving search. This method is algorithmic filtering – This is what people often end up talking about when they discuss news aggregators and other search-driven journalism.

However, there are at least five forms of filtering, and using each of them can create value differently. I think that we need to explore these other forms of filtering in trying to create online value – in the news industry as well as in other contexts.

The editor deciding what is important is expert filtering. This still is used in several contexts, such as at politico.com (discussed previously here). The expert network could be a very interesting approach to filtering new as well.

The main point here is that there is definitely still opportunity to take advantage of judgment in filtering and connecting news stories. Mechanical filtering methods (the algorithm-based approaches) appear to be dominating right now, in large part because of Google’s current gigantic footprint on the internet.

This does not mean that this is the only way to go, though. In order to create value with one of the different forms of filtering, you have to think through very carefully how you are going to do each of the aggregate, filter and connect steps. I’ve been arguing for a long time that the money in digital business models comes from filtering well, and that the firms that realise this are the ones that will do well. A business model with mechanical aggregating, and judgment-based filtering and connecting should still work. It might not be all things to all people, but then, very few successful business models are.