Blog Search Stuck in Beta

Here’s today’s column, originally published in MediaPost where there are a number of comments, including a few disagreements and some new sites mentioned which I’m trying out. Feel free to share your own thoughts here or there.

Whenever I’m talking to marketers about how to listen to their
brands, blog search engines always come up as a good starting point.
I’ve been having the same conversations for years, and with minor
exceptions, I’d do fine using the same slides I’ve used since I started.

Here are some examples of the blog search engine landscape:

Blogpulse, Icerocket, and Trendpedia:
It’s really hard to tell the difference between these blog search
engines. I like Trendpedia’s charts and interface the best, but I can’t
swear  which engine is really the best.

Technorati: You can find
useful information about blogs such as how influential they are, and
the “rising posts and stories” section is helpful, but it still has
been slow to evolve, and the one chart option available hasn’t improved
at all.

Regator: I met the team behind
this startup at Blog World Expo over the weekend. It does a decent job
at categorizing posts and pulling in multimedia, but its layout is
dizzying. Perhaps with some design fixes it’ll come around.

Alltop: This pseudo-search
engine co-founded by Guy Kawasaki is more of a blog directory. Pick a
topic and it lists relevant blogs and five recent posts. In a recent post about the latest site’s latest news, Kawasaki includes a stick figure drawing
that explains Alltop. This gets to two problems I have with Alltop.
First, if you need elaborate drawings and other documentation to
explain a search engine (including a tutorial video), then there’s
something wrong with it. Simplicity wins with search; Google never
required a tutorial. Secondly, the drawing excitedly notes at the end
that Alltop “kinda looks like a magazine rack.” I wouldn’t say that’s a
compliment, as magazine racks are horribly inefficient to find anything
besides titles you already recognize. More to the point, the Marketing category has 188 blogs listed in 63 rows. How is that an effective way to find good marketing information?

Google Blog Search: The
interface is familiar, and I like how you can easily find posts on a
specific date. The index also feels more reliable than others because
of the halo effect of the brand. Yet there’s nothing there beyond the
search engine and a few recommended blogs. With all of the resources
Google has, including Trends, Analytics, and News, Google has the
potential to offer the only blog search engine you’ll ever need.

What’s missing from these engines, broadly, is a sense of
understanding a user’s purpose. Technorati seems to do the best job
here, since it’s really a resource focused on bloggers (”tech” is in
its name after all), and features for consumers, marketers, or others
presumably aren’t as highly prioritized.

The problem for these engines generally is that people don’t wake up
thinking, “Man, I wish I could find a blog.” That happens once in a
while when people switch jobs (especially when switching industries),
but that’s not an overarching need. There are several types of
audiences, but I’ll focus on one broad category: marketers, and anyone
who’s looking for the buzz on their business, industry, and competitive
landscape. What would make blog search engines better for this audience?

1)      Better charts. Okay, so we can compare up
to three terms on a line graph (Blogpulse, Icerocket), throw in a pie
chart (Trendpedia), or use a bar chart for one term (Technorati).
That’s it?

2)      Topic analysis. Why not parse the content
to show what’s causing spikes or dips? What if there was a Farecast for
blogs, which tried to predict whether buzz would rise or fall in coming
days?

3)      Broad matching for searches. Do we really
need to search for “Barack Obama” and “Barak Obama” just to be sure
we’re getting everything? (Fun fact: Google Trends shows New Mexico
residents are most likely to search for the misspelling.)
Google handles misspellings better than others, but it falls short in
other tests, such as showing very different results for “NYC” and “New
York City.”

There’s much more that can be done, such as including blogs as just one social media channel indexed. FriendFeed
might in time emerge as the most valuable search engine because of its
advantages there. My bet’s still on Google to nail this, especially in
terms of the features that marketers need, as it has the best track
record of giving away information to bolster market share, and it’s in
the financial position to do so. Just as Google Analytics competes with
Omniture, and Google Trends for Websites competes with comScore, Google
Trends for Blogs could emerge to rival BuzzLogic, Radian6, and other
subscription-based buzz monitoring services.

That hardly means the books are shut. Heck, BuzzLogic launched an ad
network to compete with Google AdSense, so no one’s rolling over. I’d
just like to see the rivalries among blog search engines heat up so
marketers can reap all the benefits.

4 thoughts on “Blog Search Stuck in Beta

  1. I agree. Btw, I would add Twingly to your blog search engines list.
    A blog is not just text it is an entity and identity too. Technorati holds today the best blogsphere’s system of records. I think that we will see growing need for a tool helping the business to monitor and find great bloggers within different categories over time.
    I do not see any solution that meets these requirements yet!
    I wrote about it here and in few other post on my blog.
    The case for blog search engine
    http://usingit.wordpress.com/2008/08/30/the-case-for-blog-search-engine/
    Keren

  2. I agree that the current tools are too focused on the source of the data and not how this information is commonly used. The most agnostic blog search tool I’ve found so far: Google! Too bad they don’t allow for anything more than qualitative (i.e., anecdotal) analysis!

Comments are closed.