Deep Thoughts at Web 2.0

In July, we looked into the Bloglines data in a post called “Which Feeds Matter?” (which actually began life as “What Feeds Matter?” until Daniel left a comment correcting our grammar). Our definition of a feed that mattered, out of the millions of blogs and other sites with feeds out there, was one with at least one subscriber on Bloglines.

The answer was roughly 1.1 million feeds. The downside of that number was that “only” 1.1 million, out of perhaps 18 million blogs and other sites, had at least one subscriber. The upside, however, was that compared to ratings services like Nielsen, which rank only about 2500 sites per month, 1.1 million sites that “matter” to at least one person seems like a huge number.

Many of you asked for even more data, for example Greg from Findory asked how many feeds have 20 or more subscribers, etc. So as a follow-up to that post, we gave a High Order Bit at Web 2.0 conference yesterday called “What Matters?”

There were three parts:

1. Which Feeds Matter? Part 2.
2. What Matters in Search?
3. What Matters to Searchers?

Here is a link to a copy of the presentation for those who are interested.

Some highlights:

As an update to the original post, we first saw that the number of feeds with at least one subscriber had grown over the previous 4 months from 1.1 million to 1.3 million.

Next, we saw the drop-off of feeds that “matter” is significant when it comes to those with at least 20 subscribers. Jokingly, we called those feeds that “Really Matter.”

At 1000 or more subscribers, we saw that there were only 437 feeds. We gave those a more scientific term to describe their level of mattering.

Given our Web 2.0 host John Battelle’s recent bestseller The Search, we thought it would also be appropriate to look into “what matters in search.” For example, at Ask Jeeves, like other major search engines, we have grown our search index to billions of documents, and we believe that comprehensiveness – or more specifically, what we call internally “completeness” to signify that we want more pages that matter, rather than mere index growth – is one important ingredient to delivering relevant results. Across billions of searches per year, however, it is interesting to see that the average searcher rarely accesses this data. A mere 350,000 documents in our index receive 25% of all clicks into our engine in one year. A mere 6 million account for 50%.

On the revenue side, there was some interesting dialogue in the blogosphere earlier this year, with many assuming that search engines are raking it in due to their newfound ability to monetize the Long Tail of searches. Yesterday we showed that while the Long Tail certainly is valuable for advertisers and contributes to the overall value in search marketplaces like AdWords or our new Ask Jeeves Sponsored Listings, the majority of revenue in search is still made in popular searches rather than the Tail. In fact, over 70% of search revenue is made on just 30% of the searches. This shows both the stiff competition that exists for popular keywords and the upside that still exists for paid search as that competition spreads to the tail.

We’ll continue to post interesting data about blogs and search as time goes on.

Jim Lanzone