Jeff Clavier's Software Only

FOX going for Google in multi-year search deal

I just spotted on TechCrunch that Fox Interactive Media has chosen Google to power Internet search on MySpace and its other Internet properties. Per the official press release:

The agreement calls for Google to power web, vertical and site specific search for MySpace.com and the majority of Fox Interactive Media properties. Google will be the exclusive provider of text-based advertising and keyword targeted ads through its AdSense program, for inventory on Fox Interactive Media’s network. Google will also have a right of first refusal on display advertising sold through third parties on Fox Interactive Media’s network.

Whilst interesting details of the deal, such as the revenue share percentage, have not been disclosed - there is a 3+ year commitment for Google to contribute an aggregate revenue share of $900M. This deal is obviously significant as it is one of the last major US search deals that was not yet Google's. It seems to be a logical choice as Google is known to offer the best overall yield and MySpace's traffic is so large - not even mentionning other large properties like FoxSports, IGN or FoxNews that are already monetized - that Google's inventory will definitely be put to work here. The $900M figure, while “non trivial”, strikes me as low - relatively speaking. $900M over three plus years implies less than $25M a month in direct revenue to Fox. Views on that anyone ?

A derivative of this deal is that  MySpace and Google could develop, and deploy, social search at real scale if ever they felt this was an interesting experiment/trend.

Rafat provides a summary of key facts mentioned during the analysts conference call:

• Video not part of this deal, but we will have talks with Google on future opportunities.
• FoxSports.com is NOT part of the deal, due to its existing deal with Microsoft/MSN.
• This is a standard Google ad-revenue sharing deal where the majority of the revenues comes to the content holder.
• We have been planning on releasing a MySpace toolbar and we will think of integrating it with Google.
• It is an all cash deal.
• We are about to launch IGN properties in UK.
• We are about to cross 100 million people profile on MySpace.
• The deal covers every territory except two countries (not disclosing these two countries).
• We had conversations with Google competitors.
• On the remnant inventory– with which we work with 17 providers–Google will have the first right of refusal for this inventory.
• We find that the largest amount of people leave MySpace to Google…that was the most attractive part in the decision leading to the deal.
• We had been using Yahoo in some parts, and our own technology in others.
• Ross Levinsohn: I am not concerned about the relationship between AOL and Google.
• Early on we were looking fairly closely at video search, and the text search only became more important over late.

Tags: google, myspace

Gnomedex 6: Should TagJag be funded ?

Left to right: Rick Segal, me, Brad Feld 
Photo credit: Scott Beale.

Chris Pirillo had launched a few months ago a meta-search engine he named gada.be, now TagJag (see the most recent TechCrunch review here) and was wondering how he could bring scale to that business. One of the obvious solutions being external funding, he had the idea to do a VC pitch in front of the Gnomedex audience and asked VC blogger Rick Segal to lead that session. Rick in turn kindly asked buddy Brad Feld and myself to participate, and provide our candid feedback on the plan. Knowing how “candid” Rick, Brad and I are known to be, Chris’ decision to get feedback on his first ever financing pitch in public was gutsy. And the dude was clearly very nervous.

Chris did a brief demo of TagJag and explained the vision behind the product: be able to query at once a multitude of search engines in a way that is simple from a cell phone or a desktop. It also meant to be easy to use for every day users (as opposed to geeks). Finally, there was a notion to allow bloggers or site owners displaying TagJag results to get a piece of the action if the traffic they generated as monetized.

I made a few (hopefully useful) comments during the following discussion, which unfortunately was only 30 minutes long, and wanted to add a few things that I did not cover. In no particular order:

• Chris did a good job in providing a vision statement for the product – with a great deal of passion. However, he made the vision a bit confusing by mixing issues of implementation, strategy, resources and repeating how nervous he was. I explained that I prefer entrepreneurs to first give an overview of the What (what is the product/service and the value to users), the Why (why do I this and why me/my team) and the How (providing a very high level idea of an execution plan). And then dive into other considerations.
• The question at hand was whether TagJag should be funded. I pointed out – and have written many times about – the fact that there are many different ways to fund a business (cashflows from operations, cashflows from another business, personal savings and credit cards for small amounts, bank credit lines when there are revenues, and yes angel and vc financing – either as debt or equity), and there should be no reason for entrepreneurs to feel that they have to raise professional money in order to be successful.
  Brad pointed out that Chris wanted to talk about funding, but did not highlight for us what he would use that funding for. Obviously staffing up, but he only had a shallow idea of his needs (he responded 6 to 10 people and $600K – which does not really compute given where Tagjag is today). When discussing funding, entrepreneurs need to be absolutely clear as to how much is sought, what the funding is going to be used for (VCs often call that “use of proceeds”) and what milestone is going to be reached by the company thanks to that funding.
• Another important point addressed by Brad is the issue of overall scale, and potential outcome, of the business. If you are building a lifestyle business for yourself and a small team, don’t bother bringing in professional investors. These businesses can generally be bootstrapped with very little capital – personal or friends and family, and get to be cashflow positive pretty quickly. On the other end of the spectrum, you might want to build a business that will require millions of dollars to develop and operate – like a search engine for example, and it will be necessary to get funding. But in that case, the exit (aka “liquidity event”) will have to be commensurate to the amount raised from investors, and the risk they are taking in supporting you.
• Unlike the audience, we did not have a big hang up on the business model. There are clear precedents to monetize search, and I am much more interested in knowing how users and traffic are going to be acquired, and what sort of demographics will be targeted.
• I did not mention this but it is important to remember that the goal of a first pitch is to come back for a second, generally longer, meeting. So no need to try and cram everything in that one first pitch.
• The value of an aggregator like TagJag or (A9’s OpenSearch) is the relevance of the results it provides. The current implementation does not attempt to provide any sort of ranking or aggregation of individual results. Having TagJag deliver contextual and behavioral filters will be critical. Jake Luddington actually told me later today that these algorithms had partially been implemented.

In conclusion, our recommendation to Chris was to gather feedback from all Gnomedexers, and figure out how far he can push the develop of the idea, and the service, without any professional funding. Rick, Brad and I have committed to help him out figure out some of these issues, and some day, we’ll see whether Tagjag should get funded.

VC bloggers have done a good job demistifying the whole financing process, terms, requirements, etc. and entrepreneurs should absolutely take advantage of these. And they should definitely try to get in touch with VCs or Angels to get some feedback on their ideas or their plans. It might not work with everyone you contact all the time (professional investors tend to be extremely busy people) and it might sometime take weeks to get some feedback (this is a public apology to all the people who have emailed me in the recent weeks and with whom I have not followed up yet), but through your contacts, or at a conference, you should be able to get a few minutes of an investor’s time.

Kevin O'Keefe also covered the session: TagJag : Gnomedex wrapping up with Pirillo pitching tag search engine.

Update: here are the landing page of the podcast, and the MP3 file.

Tags: chrispirillo, tagjag, gnomedex, bradfeld, ricksegal

Wikio opens its door to 5,000 beta users

Not to be mixed up with Wikia, Jimmy Wales’ new company, Wikio is the new startup of another Internet entrepreneur, Pierre Chappaz. Pierre is famous in France and Europe for having founded and developed Kelkoo – a European shopping comparison engine – to a half-billion USD exit to Yahoo a couple of years ago. After taking a year off, Pierre joined Index Ventures (the new star VC firm in Europe) as an EIR, and started Wikio – which has now a dozen developers and editors.

What is Wikio ? Topix.net meets Digg meets Memeorandum with a zest of Wink – sort of. And all in French – for now. The private beta-site just opened last night to 5,000 members of the “Francosphere”, but has an ambition to launch throughout Europe. No decision has been made regarding a US version according to Pierre, who has bootstrapped the development to date. Like so many consumer Internet plays, the idea of Wikio came about because existing tools: Google News, Technorati, etc. did not provide homogeneous search across blogs and traditional news, offering correlation between stories and the ability for users to comment and contribute.

Wikio today aggregates stories from 10,000 sources (in French), eventually growing to 50,000 across European languages. Blogs and traditional news outlets are manually selected, reviewed and classified by a team of editors based on quality and content. Upon reception of a story, the system is extracting semantic elements that will be used to correlate stories automatically, as well as generate topical tags. At least a portion of the tags are organized in a comprehensive taxonomy – not clear how this will scale, but the blend of folksonomy/taxonomy seems to work well.

The user interface provides a powerful navigation: top categories (Economy, Business, Technology, etc.), tags, most recent, most relevant to a topic and most popular. The popularity comes from a Digg-like voting system that allows readers to vote for a given story. Searching for a given keyword either delivers a straight keyword search or if the keyword matches an existing category or tag, displays the “landing page” for that keyword. In the latter case, it is also possible to restrict the scope of a search to a level of the taxonomy.

Beyond voting, users can comment on a story as well as contribute one through a standard blog-like interface.

I have to play more with it, but I find the service full of good ideas – bringing the strength of many services I use every day in a single one.

There are questions about how different Wikio is from Google News, Technorati, Digg, Wink, etc. Here are a few thoughts:

• First, as I just wrote, it integrates feautures from all the above, making it more powerful.
• Second, and this is more a business issue, Wikio might well decide to remain in Europe and cater to local language requirements that are traditionally not always well served by US companies.
• Third, Wikio is an interesting mix of editorial selection, user generated voting and content, on top of a unique blend between taxonomies and folksonomies.

You can register for access to the beta site here, and read more about the service on the Wikio blog. And if you really want to take a look, leave a comment, I am sure that Pierre will arrange a few more beta accesses.

Almost forgot: I have no financial interest in Wikio, but I am delighted to write about a French startup for once.

More:

• Pictures (clicking on them will blow them up): the first one shows the wikio home page, the second shows Web 2.0 related topics, and the third shows the publication interface.
• And of course, TechCrunch has an extensive review – en francais dans le texte :-)
• Actually, Ouriel has also pointed me to the English version of his post (thanks).

Tags: wikio

Yahoo gives up to Google on Search ? You really believed that ?

I was alerted by Thomas Hawk about the latest “news” building up in the blogosphere, and as such the top news on Memeorandum: Yahoo! gives up quest for search dominance. My initial reaction was that this must be one of these quotes being taken out of a specific context, and the sensational aspect of the headline is leading everyone to pile in. To cut the commentary short: Bullshit. No way.

Google has by far the largest market share in the search market – in the US and Europe, and it will be difficult for Yahoo to catch up on pure algorithmic search. Asia is a completely different matter, and is fair game to both players – just listen to the end of Om and Niall’s most recent Podsessions on that very topic. It is also worth remembering that we are just at the beginning of the search game: alternate search mechanism are being developed (as we discussed during our last session of the Search SIG), complex searches don’t work (try a search on “fridge” “same day delivery” as I did after our fridge melted on Sunday – I ended up calling retailers in the area b/c search results were so off base), etc. There is so much to be done…

I am not defending Yahoo here (they are big enough to do it themselves :-), or downplaying Google (my default search engine), but merely wishing that we bloggers think about the implications and/or the likelihood of what is being reported before adding noise to the signal. Too much trigger happiness and not enough analysis in my opinion. Dave Taylor is apparently in the same boat:  No, bloggers, Yahoo isn't “conceding search to Google”.

More later. Got to run to a meeting – with MSN Search.

Get ready to spot some RoofSense ads

The MIT Advertising Lab points to this picture of a building rooftop covered by the branding of the business occupying it (in the example: Target). It turns out, reading the comments related to the post, that this Target store is located next to O’Hare airport (Chicago, IL) – the busiest airport in the country, on the path of one of the runways.

The point though is that this kind of practice might become a new advertising mechanism, sort of the billboard-equivalent of satellite photos. With GoogleMaps and MSN VirtualEarth providing high-definition photos of cities, one could figure out some way of displaying advertising/branding messages on rooftops next to popular monuments or coordinates. As to whether this would overlap Google or MSN’s own advertising plans is another story.

Create any kind of new “real estate”, advertising will be one of the first leveraging it.

[via Search Engine Roundtable]

A few data points regarding the Travel meta-search market

if you are interested in tracking the shopping/comparison engines market, I recommend the work of Brian Smith over at ComparisonEngines.com, where he mixes news with interesting analysis of that space. I discovered this blog when Brian wrote about client Kaboodle.

Brian recently commented on the move of Yahoo Shopping's Rob Solomon to SideStep, a mature player in the meta-search engine market, and mentioned a number of interesting facts concerning the online travel space - a $60B market:

• As for travel search (or travel meta-search engines), no one is releasing numbers, and total bookings are tiny compared to those from Online Travel Angencies (OTAs). However, there are a lot of factors which seem to be pointing to a tipping point in the industry: AOL driving traffic to Pinpoint Travel, Yahoo putting FareChase on its main travel page (as opposed to keeping it separate from Travelocity), Google working on travel search, established travel suppliers like Hilton Hotels and American Airlines signing deals with the travel search engines, Amazon choosing to partner with SideStep as opposed to an OTA, DCA3 carrier agreements up for renewal at the same time travel providers are seeing success with direct bookings, Mobissimo (and others) searching Orbitz, Kayak working on cool Google-like aps and an impressive travel only PPC engine, etc.
• SideStep is the leader in the meta-travel search arena, but the market itself is tiny in comparison to the online travel agencies (OTAs) like Expedia and Travelocity.
• Word of mouth adoption will continue to be important for the company, but SideStep needs to step up internet marketing and more importantly, business development activities. When I say internet marketing, I think there is a place for PPC marketing on Adwords and Yahoo Search Marketing, but the problem is that the OTAs make more money on every sale/lead so they are going to be able to outspend SideStep, Mobissimo, and Kayak.

Travel is obviously a huge online market, which explains why so many players have been able to sustain themselves with very limited differentiation - at least from the standpoint of the consumer. It will be interesting to see if a consolidation of some sort happens in 2006 - similar to 2005 that saw shopping.com, ShopZilla and more recently PriceGrabber all being acquired.

Search Consolidation Act #1: Truveo goes to AOL

This post initiates a new consolidation series - about the Search market, similar to the one I have been writing about Social Media M&A. With the increasing number of search players, this consolidation is to be expected as large Internet companies (the famous GEMAYANI) look at acquiring assets to strengthen their own market position.

As it is often the case, I did not expect that Act #1 would actually involve one of my own companies: Truveo. Truveo, the little video search engine that could, had developed over the past couple of years pieces of unique technology that allowed the company to better any other search engine on discovering videos on the web. The company was started early 2004 by my friends Tim Tuttle and Adam Beguelin, who bootstrapped the venture for a year before raising a small round of financing that I was very lucky to get involved as an angel investor.

After building the core technology for about 18 months - including their core asset, the "visual crawler" - the team deployed their crawling infrastructure, and started building their index of videos found on the Web - leading to a successful launch in September 2005 of a destination site (Truveo.com) that would be followed by distribution deals enabling other search engines to leverage Truveo's index. It just so happens that there was such an interest around the team and the technology they had built that M&A discussions replaced partnership discussions - leading to the AOL acquisition is announcing today. The deal actually closed on Dec 21st  (and investors got nice Christmas gifts in their shoes bank accounts two days later). And no, ladies and gentlemen, I will not give any clues as to the size, magnitude or weight of those gifts since AOL has elected not to disclose the financial terms of the transaction (which is too bad).

The rationale from AOL's standpoint is explained in this quote, found in the press release:

“Consumers across the web have eagerly anticipated a video search engine that offers relevant, up-to-the-minute, premium video results,” said Jonathan F. Miller, Chairman & CEO of America Online, Inc. “The addition of Truveo’s video search technology, video assets, and experienced talent complements the overall video experience on AOL and takes it to new heights. Truveo is a hidden gem in the video marketplace, offering AOL members and users of AOL.com a vast, new array of tools and an extensive collection of updated, hard-to-find video assets. The Truveo acquisition takes AOL even farther down the path of being the premier destination on the internet for video and search.”

The press will most likely ask Tim and Adam why they decided to sell “so early” as things were genuinely going very well: the company had generated a lot of business interest in the US market and overseas, and could have gone the execution route - and I have no doubt that we would have been successful. At the end of the day, and I am only commenting on my own account here, I believe that AOL was clever enough to offer compelling arguments in favor of the acquisition, and add a unique asset to its video arsenal.

So congratulations to Tim and Adam, and the team, for having built a truly innovative company, and executing almost “by the book” - including this exit (shoot - I can't use any qualifier here). It was a pleasure working with you all, even if I could not get them to start a blog (though it is not too late). And best of luck for the integration within AOL.

Oh yeah, and anybody announcing the deal before 6am PT has broken the embargo :-P.

I just saw the Reuters brief story on the acquisition and was amused by that statement:

Financial terms of the deal were not provided.
The deal was the largest purchase by the Time Warner Inc. division in 2005 and smaller than the $435 million purchase of Advertising.com in 2004, an AOL spokesman said.

One can only wonder why the actual number is not announced, but anyway we know it is the largest of the year for AOL , and it is less than the $435M we got for Advertising.com (the fund I was GP of before starting SoftTech VC was an investor in that company). Staci Kramer from PaidContent was the first to call an actual number: $50M (note that I am not confirming anything).

Lots of coverage of the deal (as usual, Memeorandum lays it out for us).

Tags: AOL, truveo

SDForum Search SIG - Search Different: Tagging, Social Bookmarking and Sharing - Jan 10th @ Yahoo - Featuring Del.icio.us, Digg, Kaboodle & Wink

As already announced before the holidays, we are looking forward to a great event for the first session of 2006 of the SDForum Search SIG, and hope that you will be able to join us.

The session will focus on “alternate” search mechanisms that have been developed over the past 12 months such as bookmarking, tagging, rating, sharing and any combination thereof – generally in a social context. These concepts have grown to being at the core of many Web 2.0 companies, and offer a different way of finding or discovering “stuff” online – as a complement or alternative to traditional search engines.

As usual with Search SIG events, we will have a panel moderated by the host of the session, followed by demonstrations from presenting companies. Our host for the event will be Charlene Li, from Forrester Research and will be joined by 4 founders of exciting startups:

• Joshua Schachter - Del.icio.us (recently acquired by Yahoo)
• Kevin Rose - Digg
• Manish Chandra - Kaboodle
• Michael Tanne - Wink

This session will take place on Tue Jan 10th at Yahoo’s HQ (701 First Avenue, Building C, Room 4&5 in Sunnyvale, CA), and will promptly start at 7:15PM (registration from 6:45PM on).

As to the agenda:

6:45-7:15pm - Registration / Food & Drink
7:15–7:20pm - A few words about the Search SIG
7:20-8:20pm - Act I: State of the Art of “Searching Different” through a panel discussion led by our moderator, Charlene Li
8:20-8:30pm - Intermission: Search Networking & Geeking Out
8:30-9:00pm - Act II: Product Demos from Del.icio.us, Digg, Kaboodle and Wink – and Q&A with the audience
9:00-9:15pm - Open Mike Geek: Announcements (30 seconds of fame & fortune)
9:16pm - You Don’t Have To Go Home, But You Can’t Stay Here

Attendance is free for SDForum members, and charge is $15 for non-members – paid at the door. You can register for the event on the SDForum website, where you will also find the latest information about the session.

Note to Yahoo employees: the session is free of charge for you, but the pre-registration system does not handle your specific case. Just identify yourself at the door on the day of the event.

And remember, there is a distribution list for Search SIG announcements – if you want to join, send an email to sdforum_searchsig-subscribe@yahoogroups.com.

Tags: delicious, digg, kaboodle, searchsig, wink

SIG Session on Alternate Search: tagging, bookmarking and social sharing - Jan 10, 2006 at Yahoo HQ

Before everyone disappears on holidays, here is a quick pre-announcement regarding the January session of the Search SIG, which should take place on Jan 10th (though the date is not 100% confirmed) and will be hosted by our friends from Yahoo at their headquarter in Sunnyvale.

The theme of the session will be Alternate Search, and will cover the mechanisms developed over the past 12 months to create new ways of indexing and discovering information: tagging, bookmarking,... shared amongst friends or communities.

Charlene Li has kindly agreed to be the moderator of the event, that will involve Wink's Michael Tanne, Kaboodle's Manish Chandra and a few other executives who have not confirmed yet.

Tags: searchsig

Search + Tags + People = Wink

Michael Tanne, the CEO of Wink, briefed me a couple of days ago on the forthcoming public beta that became available tonight around midnight Pacific Time. Wink is a search engine that allows users to bookmark and tag search results sourced from other popular bookmarking sites like del.icio.us, Digg, Slashdot, etc. Wink has an implementation of the concept of leveraging user's gesture of attention - in their case tagging - to  thereby altering the order in which results are going to be displayed, and somehow rank results internally, an approach already implemented by services like Digg or Tailrank. Wink also displays Google results  below its own, which  provides an easy mean to compare  Google and Wink on certain queries.

There are a couple of neat features that just got released in this version:

• It is possible to seed one's tag history by importing the content of a del.icio.us or MyWeb archive, which is a great idea. Even better, wink's and del.icio.us' caches can be kept in sync. This is a great idea.
  Update: TDavid has a gentle go at me on the "neatness" of the functionality I have listed, so I am expanding on this one (from a comment I left on his post):

  The reason why the del.icio.us import is compelling is that I find this whole idea of using my tags to drive the relevance of my search results interesting. But I don't want to be forced to use a wink tagging system to do that - and that's why I don't use MyWeb even if MyYahoo is my home page. I want to use the tagging system/framework(s) I like, and then have Wink do the work to integrate them.
  Furthermore,there aren't that many tags out there and it would take a long time for Wink to get enough tagged information to do something useful/meaningful on the relevance front.
  Finally on the synchronization of tag caches, why not ? If for some reason I find a search result through Wink that I want to tag, knowing that my tags will also end up in del.icio.us is a matter of convenience. Hence the neat.

• Wink now offers an Answer feature, in the form of a wiki page that can be edited by the user community. It actually uses Wikipedia content as initial content when a definition is available for a given query. For example this ego search picked up my Wikipedia entry, whereas I manually entered this one.

More about Wink's functionality can be found on the Wink blog announcing Obi-wan OB1, and of course, TechCrunch was the first to cover the news.

Wink's business model will most likely be based on some form of advertising, and the usual AdWords are already on the site. What will be interesting to monitor once a critical mass of tags have been amassed is whether Wink has a better yield than “plain Google” thanks to the increased relevance brought by tagging.

Actually, Michael will be one of the participants of the January Search SIG that will cover “Alternate search: tagging, bookmarking, social sharing”.

Tags: wink