Tomorrow AOL will announce the acquisition of San Francisco-based Sphere, a blog content engine that launched in 2006. The price is not being disclosed, but sources are suggesting it’s in the $25 million range, or possibly a little more. More details from Om Malik

When Sphere first launched as a blog search engine they were already late to the blog search game. Technorati and others had been around for some time already, and even Google Blog Search was nearly eight months old. Sphere had some nice features, but it was in a tough and competitive space.

But CEO Tony Conrad, a former venture capitalist, quickly adapted to the changing market and focused on delivering blog results relevant to content delivered by big news and content sites. Time was the first to go live with “Sphere It” links, and most of the big news sites followed over time. In July 2007 we noted that they had very quietly completed a transformation into a “related content” engine.

Sphere lands in Bill Wilson’s organization, the EVP of Programming at AOL. His division controls AOL’s content properties (Entertainment, Finance, Weblogs, etc.). In a phone call today, Wilson told me he doesn’t intend to change Sphere’s approach or brand. They are growing a number of micro/niche brands, he said, and leveraging what he calls “passion points” of small but passionate audiences. Sphere fits right into that by showing relevant content to users, and getting AOL content in front of more users.

Congratulations to Conrad on the aquisition, as well as the rest of the Sphere team (Martin Remy, Steve Nieker, Toni Schneider, Mike Garfias, Alex Bendig, Andy Cabell, Anne Dorman, Jeff Yolen, Adam Embick, Josh Guttman, Kevin Cowan, Sven Henderson, Troy Vitullo and Michael Harzheim). Sphere had raised $3.5 million in venture capital over two rounds.

At first glance, blog search as a category is oversaturated. Ok, at second glance, too. Not only did Google enter the market directly in late 2005, they’ve also increased the rate that they index blogs and other regularly updated sites for core Google search. TechCrunch, for example, is now indexed multiple times per day by Google, and new posts are often available in a normal Google search within minutes of posting. Most people today say the best blog search engine is, simply, Google.com.

And there are many competitors. The Comscore chart below shows the relative traffic of the major ones – Technorati, Google Blog Search, Ask Blog Search, Sphere and IceRocket. Feedster is gone, although there are additional smaller engines like Zuula and Blogdigger as well. Every one of those companies is U.S. based (note that Paris-based Wikio has blog search as well as a Digg-like service).

Now Europe will have it’s own blog search engine – Twingly. I met Martin Källström, the company’s CEO, at the DLD conference in Munich earlier this week. Their focus, he says, will be to have a spam-free engine (something none of the others can claim) at the cost of inclusiveness. And at least at first, the engine will be focused on European blogs. Twingly’s search engine hasn’t launched yet, although I do have a screen shot of what the home page will eventually look like:

Twingly already has a product – a nifty screen saver that shows blog posts on a world map as they are written. The new search engine will use some of the back end technology they’ve developed for the screen saver – mainly their ping server (see here for our overview of what ping servers are) and existing index of blogs.

The search engine will be different from others, Källström says, in that it will be almost 100% spam free. How are they doing that? Instead of trying to index every blog in existence and then removing spam via black lists and other methods, they are limiting the blogs they monitor to those that are proven to be legitimate. They started with a small list of known blogs, and then spidered out from there based on links to other blogs. The assumption, which is fairly sound, is that good/real blogs will not link to spam blogs. The end result is a white list of real blogs that are indexed – everything else is ignored.

Källström says that, in addition to the consumer-facing search engine, they’ll partner with large content news sites to show blog posts related to news content. This is something both Sphere and Technorati have had success with in the past, and the company can do revenue-sharing deals on additional page views. Content providers like it because it incentivizes blogs to link to their content (to get a link back). Twingly may not be able to compete with Sphere and Technorati in getting U.S. based partners, but he says he already has some deals with large European publishers completed.

The company has raised €1 million in a July 2007 round of financing from Servisen. They have seven employees. Look for a launch of their search engine in the next month or two.

Blog search engine Sphere has kept a relatively quiet profile recently. After a private beta launch in late 2005 and a big-press full launch in the spring of 2006, they’ve kept mostly to themselves.

Instead of focusing on building a better blog search engine (an area that Google now dominates after a recent Technorati refocus), Sphere has spent its time developing technology that automatically finds blog posts related to whatever a user is looking at currently. The product, called SphereIt, has been added to around 20,000 blogs (including TechCrunch, see the end of any blog post) and gives users additional reading material on the topic of the article or post. It is one of the top 50 plugins for Wordpress-based blogs.

But the company, led by CEO Tony Conrad, isn’t just attacking the market from a bottom up approach. They have also been closing deals with some of the biggest news sites in the U.S. Deals with The Wall Street Journal, CNN, The New York Times (Tech & Science sections), AOL News, TIME, Dow Jones Market Watch, CBS News, AOL Entertainment, NBC Access Hollywood, Media General Affiliates, AOL Sports, ZDNet Blogs, DWELL, About.com and others were all announced recently on the Sphere blog.

These news sites and 20,000 blogs all have Sphere results included on story pages – see this story on CNN for an example, or the screen shot above for the WSJ implementation. That’s well over 1 billion page views per month with Sphere links and results.

Competitor Technorati has pursued deals with these and other sites to promote their results as well. But Technorati only returns results linking to that particular story, meaning most relevant content is ignored. Sphere instead analyzes indexed blogs and performs a semantic analysis on the text to determine relevance. The results are strikingly better – and the fact that the WSJ dropped Technorati and added Sphere attests to that.

Sphere doesn’t pay any of these sites to include their results. Instead, they will serve advertising in the results and share the revenue with the partners.

The company has done a lot with just under $4 million in venture capital and 9 employees. They’re technology seems to be proving valuable to partners…which makes them perfect acquisition bait down the road. I don’t expect to see Sphere in the deadpool any time soon.

When blog search engine Sphere launched in May 2006, it included a unique feature that discovered, on the fly, stories related to what you were reading regardless of whether or not the two stories were hyperlinked.

The feature, called “Sphere It,” has grown in popularity and has helped Sphere get itself embedded in top blogs and news sites. We include a Sphere It link at the bottom of each post, which pulls up a window where other blog stories that talk about related issues are shown. Time.com and other major news sites have done the same. Today, a substantial portion of Sphere’s total traffic comes from these partner sites using the feature to generate more content (and page views) for readers.

Tonight Sphere relaunched their home page to reflect the usefulness of this feature. The main area of the site is broken down into four columns. On the left are major topics, like Top News, U.S. News, Entertainment, Sports, Technology, etc. Click on any topic and the second column populates with recent news items from Sphere partners (sites like ours, Time.com and others that include the Sphere It functionality) that has generated a lot of buzz, which is calculated based on page views for the item (against an average for the site) and other factors the company isn’t disclosing (but which probably include an anlysis of the extent to which other sites are writing about similar things).

Click on any news item and it is pulled up in column three. In column four, related news items are shown.

You can keep clicking on a topic infinitely. Click on a column four story – it moves left, and column four shows news items related to that story.

Like TechMeme, the new Sphere site can become a place that people check frequently to see what news is breaking in the blogosphere and mainstream media, and see other content about that topic. It’s different than TechMeme in that Sphere doesn’t require links between articles.

News sites that want to be included in Sphere It must add the feature to their sites. The Sphere home page has a link to do that.

Sphere has raised just $4 million to date from True Ventures, Trident Capital and a number of angel investors. The company has eight employees.

The annual Web 2.0 Summit kicked off today at the Palace Hotel in San Francisco. The conference Summit, which has been sold out for months, is noticeably larger than last year and hundreds of people are milling about, seeing and being seen.

The highlight of last year’s conference for me was LaunchPad, where thirteen young startups showed their stuff to the audience. See our coverage from last year here and here. Many of those companies are doing very well. Only one, Pubsub, has entered the TechCrunch DeadPool.

LaunchPad this year was perhaps even more competitive than last year. Over 200 companies applied to present at the conference. Only thirteen were accepted, and each had five minutes to demo their product to the crowd. We have a summary of what each announced below.

Read More

San Francisco-based Sphere’s only been around for three months, but it’s already locked down two important deals. Both deals leverage its “Sphere It” technology, which performs a semantic analysis on the text within the page being searched and returns blog results that it finds relevant to the article.

In May we wrote about Sphere’s deal with Time.com. Next week, Sphere will announce a deal with the New York Times subsidiary About.com to embed the core Sphere It technology within About.com articles.

Unlike the Time.com deal, About.com will only use the technology to sort through it’s own content. The results can be seen under the main text of articles, under the heading “Related Articles.” See here for an example.

The two companies have been in an extended year long paid test of the technology (using it long before the core Sphere search engine launched). According to Sphere CEO Tony Conrad, the test performed so far above expectations that About.com has agreed to a long term deal to license the technology.

Competitor Technorati has its own deals with major partnerships like the Wall Street Journal Online, Washington Post and Newsweek. Those deals, however, point to blogs posting on a specific article, whereas Sphere It focuses on the content itself, finding similar articles regardless of whether or not they’ve linked to the original source. See this post for a longer description of the difference between what the two companies offer (frankly, both are useful, although Sphere It is better for newer content that has few links in).

San Francisco-based blog search engine Sphere is less than a month old, and already has links to its “Sphere It” functionality embedded in a number of time.com articles (example). Tony Conrad, Sphere’s CEO, says that they are currently testing the feature with Time.com.

The links are prominently placed below the headlines of articles and link directly to Sphere blog search results related to the topic. I wrote about Sphere It in my initial profile of Sphere – under “more tools” near the end. Sphere It is a Sphere feature that allows users to find relevant blog content from any URL. The easiest way to use it is to install the Sphere It bookmarklet into the browser. Click it while on any web page and relevant Sphere blog search results will be brought up.

Unlike Technorati’s “Technorati This” feature, which shows blog entries that link to the URL being searched, Sphere It doesn’t report links. Rather, it does a semantic analysis on the text within the page being searched and returns blog results that it finds relevant to the article. For more on the difference between Sphere It and Technorati This, see Sphere CEO Tony Conrad’s blog post here.

To see an example, the Sphere It results for the above article, on General Michael Hayden’s nomination to head up the CIA, returns these Sphere blog search results. A quick perusal shows what I consider to be very relevant blog content. The speed of the search, which is done in real time, is also impressive.

While this type of exposure is excellent for new startup Sphere, it’s also great for the blogosphere. I would imagine the average Time.com reader is not fully aware of what blogs have to offer in terms of breaking news and (often) intelligent commentary. This should drive more mainstream Internet users to the blogosphere over time.

The Sphere It bookmarklet is available here.

New blog search engine Sphere launched just moments ago and has also announced a $3.75 million round of venture financing. In addition to covering the launch of Sphere here, we have a podcast interview with CEO Tony Conrad and advisor Toni Schneider over at TalkCrunch.

Sphere, which follows in the footsteps of previous blog search engines like Feedster, Technorati and IceRocket, as well as offerings from Google and Yahoo, is doing things quite a bit differently than its predecessors (and its evolved dramatically since our first look at it back in October).

Sphere’s design, by the way, was created by Adaptive Path. See their essay on the project here.

The site is segmented into three main areas: blog search results, featured blogs, and related media.

Blog Search Results

Sphere search results can be viewed by date, relevance or a combination of both. Unlike Technorati, which determines a blog’s relevance based on the total number of unique links into that blog, Sphere is taking an algorithmic approach. For Sphere, “relevance” is based on three key factors: links in/out of blog; meta data around the blog (average length of posts, post frequency, etc; and a semantic analysis of the posts themselves). In our tests Sphere blog relevance is very good. We’ll do a more in depth review and comparison at a later date.

Another feature is a “custom date search”. In addition to preset date selections, if you do a custom date search by selecting it in the drop down box, you’ll get results just for that date range and you’ll also see a day by day breakdown showing results per day with the included term.

Each search result has a link to a blog profile that includes basic blog information (links in and out, average post frequency and length and additional information). In the future an extended profile for each blog will be available that will include information from the blogger as well, such as a photo, a zip code for geotagging, and topics the blog covers.

Featured Results

The most relevant blogs for thousands of search terms are listed in the “featured results” area.

If you don’t see search results for what you are looking for, click over the the most highly rated blogs for your search term to research further. These results are based on the same algorithmic analysis as blog search.

Related Media

Want to go beyond blogs for your research? Click on “related media” and see related pictures, news, books and podcasts relevant to your query.

More Tools

As great as the basic search platform is, what I like best about Sphere is in the Tools area. Install the “Sphere It” bookmarklet and click it whenever you are reading something that you’d like more information on. Sphere will analyze the page in real time and present blog search results that are relevant to that topic. It’s important to note that this is not a search to find blogs linking into that page you are viewing; rather you are finding fresh blog content that is related to the subject matter of what you are reading. I’ve tested this and find it extremely useful.

Congratulations to the Sphere team for getting this launched, and taking blog seach another step or two forward.

Company: Sphere
Launched: Private Beta Launch this Weekend
Status: Funded by Doug MacKenzie, Kevin Compton, Phil Black, Will Hearst, David Mahoney, Vince Vannelli and Mike Winton
Previous Post: October 8, 2005
Location: Palo Alto

I met with Tony Conrad, one of the founders of Sphere, today at our office (ok, its a house, but it’s also an office – just ask the IRS). Sphere is a new blog search engine that quite frankly blows everything, and I mean everything, I’ve seen out of the water in terms of relevance (and, by the way, design…Adaptive Path was involved).

The timing for this meeting was perfect – right on the heels of Yahoo’s new blog search engine, and Robert Scobles already-classic post “The race to time-based and blog search“.

Tony, along with co-founders Steve Nieker and Martin Remy, and advisors Matt Mullenweg, Mary Hodder and Toni Schnieder, have created Sphere in a very short period of time and for “less than $200k”.

I saw the live site and was allowed to bang on it as much as I liked. I did, and came away very impressed.

Relevance in blog search is very difficult. Google-type PageRank analysis, which looks at incoming links to a piece of content, simply doesn’t work because new content doesn’t have much in the way of links. Until now, no one has come up with a way to properly sort blog posts by relevance, and the general default way of showing results is “reverse-chrono”, which simply puts the newest stuff at the top.

Sphere appears to have solved the problem, or at least taken big steps in the right direction. Their approach involves three key algorithms – an analysis of links into and out of a blog, an analysis of metadata around a post (links, post frequency, length of posts, etc.), and something Tony calls their “secret sauce”, which is content semantic analysis to filter out spam and to understand what a blog post is talking about.

Result sets show only two posts per blog on the first page, so no one blog can dominate a category. The result set has auto-generated profiles of blogs that include recent links in and out of the blog, average posts per week, average words per post and a link to a full page profile that can be edited by the blog author.

To the right of the main result set are blog (as opposed to post) links that are relevant to the query, and something they call the “Daily Sphere”, which is links to relevant non-blog news stories. I imagine they may add additional content in the side area as well.

The results page integrates both ajax and flash features in an intelligent way (it’s not there just for show).

While Sphere has been indexing blogs since January of 2003, their index only shows results for the previous four months. They will lengthen this window as they scale up operations.

Sphere has tens of thousands of beta requests and is opening up to 100 lucky people this weekend. Sign up for the beta here.

Om Malik writes about Sphere, a new blog search service that appears to be getting ready to launch.

Sphere is taking a crack at building a more relevant blog search engine. Traditional link analysis just doesn’t work with blog posts because new posts don’t have time to gather links. Instead, Sphere seems to be be trying to determine blog authority on a given subject area, and determine new authoritative blogs based on who those blogs point to.

From Sphere’s Learn More page:

What makes us better than existing blog search sites?

It starts with relevant results and fast performance. Our new relevance-based algorithm discovers new blog posts as they’re created, indexes them within minutes of being published, applies rich semantic analysis and makes them searchable by relevance or time. Plus, we’ve got a few fun, helpful features that we think make for a richer user experience.

Of course, there is no way to tell if Sphere has cracked the blog search relevance problem until they launch. I’m looking forward to finding out.

Thanks Richard for pointing me to this.