ITunes 8: The Genius in the Box, Picassa and VideoSurf

Ed: Genius = Social Graph of Songs; Picassa and VideoSurf = search based on image recognition. Emerging theme is simplify search for multi-media like photos, videos, and songs.

ITunes 8: The Genius in the Box

Music discovery services are definitely a hot topic right now, with Pandora, Last.fm, imeem, and others vying for users. Yesterday, Apple joined the fray when it released iTunes 8 and its ‘Genius’ recommendation engine. After examining your iTunes library, iTunes uploads data about your library to Apple’s servers and returns back a set of information about how the songs in your library correlate to each other. Based on this, iTunes can now build playlists of similar songs and display shopping recommendations.

How Does it Work?

As is typical for Apple, the company is not exactly transparent when it comes to describing how the ‘Genius’ feature actually works. It looks as if Apple compares your music selection to that of other users and then builds its recommendations based on this. We assume that iTunes looks at data about play and skip counts, beats per minute (which is available for all songs in the iTunes store), ratings, and playlists…

What about Last.fm and Pandora?

As Last.fm co-founder Marting Stiksel pointed out in an interview with Wired’s Eliot Van Buskirk, the ‘Genius’ feature basically validates what other music recommendation services have been doing for a long time.

It’s also important to point out that a lot of other music recommendation services have strong, built-in social networking functions. Apple, even though it now has information about the listening habits of a large chunk of its users, does nothing to connect these users. One neat function, for example, would be for iTunes to show playlists from other users that have a certain songs in it. For now, though, it doesn’t seem as if Apple is interested in adding these social aspects to iTunes anytime soon…

TC50: Scoble Declares That VideoSurf “Doesn’t Suck”

Video search is an unsolved problem. VideoSurf applies hardcore computer vision technology to this problem and finds relevant results beyond what may already be available to text-based search methods. In the demo at TechCrunch50, the startup showed how you might want to search for a scene in the show Entourage. You can drill down to the show, and then are presented with thumbnails of all of the characters, left to right at the top of the screen by level of importance. By clicking on a character, you get all the scenes in which that character appears, as well as related scenes.

A search starts with a keyword or normal guided navigation, but quickly becomes driven visually. Every time you click on a thumbnail, based on the visual cues, you are taken to exactly the right moment in the scene. Other thumbnails of related video snippets also appear. VideoSurf automatically adds tags and metadata to each and every frame of video. That is what is driving the search results. Once you click on the scene you want, you can watch it. Or, if you want to share a specific moment, you can set some time sliders and send exactly the scene you want off to your friends and family.

Robert Scoble, one of the TC50 judges, summed up his reaction (and that of much of the audience): “Finally, a Website that doesn’t suck.”

Panel Q&A:

Bradley Horowitz: I have some experience in the space. I studied computer vision at MIT. If you are looking for a Mentos video, you don’t need a Mentos detector. People capture that. I didn’t see that social [element]. Are you taking a hypertechnical approach to solving video search?

CEO Lior Delgo: “People are not going to go and tag every frame in the video. That is what we are doing, going frame by frame and creating metadata that does not exist. You can ask everyone in the audience to tag each frame, you won’t be able to do that? This technology has never been done before. This is the first time it is working.”

Joi Ito: “When you say look into the video, how are you doing it, how accurate is it?”

CTO Dr. Eitan Sharon: “We go everywhere from detecting scenes, get the unique and important moments, also extract the people. So the same person is grouped together, and all the appearances will be presented together. It has face recognition.”

Ito: “Obviously, it could also be used for surveillance. Can I look for “Scoble sex,”? What will I find?”

Michael Arrington: “Zero results.”

Scoble: “How long does it take to process and index each video?”

Sharon: “Much faster than real time.”

Scoble: “Can this be done for streaming video?”

Sharon: “We can process it in real-time as well.”

Sheryl Sandberg. “It is based on being able to understand the data in the first place. The reason AdSense worked is because there is a deep understanding of the content. What exists for text-based content does not exist for anything else. If you think about actually being able to metatag video, this is exciting. I also think that starting with a consumer Website is the right start

Horowitz: “If you look at PageRank, we look at link flux. The context is important, where does this exist in the Web. You want to go beyond just drilling down into the data. I don’t see that here.”

Arrington: “It would be nice if a computer could tell you what’s inside a picture.”

Horowitz: “Did you see Picassa? We launched face recognition last week.”

Delgo: “The information we are able to collect is way more relevant.”

NEWS: Intel Mash Maker: Mash-ups for the masses

Intel Mash Maker: Mash-ups for the masses

from CNET News.com | Tech news blog by Martin LaMonica

Intel wants to make the whole Web editable, just like a single Wikipedia page.

The chip giant on Tuesday will make a beta available of Intel Mash Maker, a free browser extension that allows users to modify Web pages and combine information from different sources. Its first beta works with Firefox 3 and Internet Explorer 7, though at this point the features are far more mature in Firefox, Intel said.

The product, which originated in Intel’s research labs, is similar to existing mash-up tools like Yahoo Pipes and Microsoft Popfly in that it has a graphical design tool.

Intel Mash Maker suggests customizations and widgets. (Credit: Intel)

What’s different is that the actual mashing up of information on Intel Mash Maker happens on the client, rather than the server. So instead of making a different Web application to, say, plot real estate listings on Google Maps, Intel Mash Maker lets people add a widget that adds visualization to the real estate listing site….

Intel’s Mash Maker: Mashups for the Rest of Us

from PC World: Latest Technology News

With Intel’s new browser extensions, you don’t have to be a programmer to mix and match data from multiple Web sources to your liking.

Intel Tries Hand At Consumer Mashups

from TechCrunch by Mark Hendrickson

Intel is releasing into public beta today a new “experimental” product from its labs called Mash Maker. It’s a browser plugin – most functional with Firefox but also available for Internet Explorer – that lets end user create their own mashups on top of existing websites.

The idea of a mashup, while very central to the movement we call Web 2.0, has always struck me as an overly techie concept. And Intel’s Mash Maker doesn’t do much to change that, even though it tries to bring mashups more mainstream.

What’s a mashup? Simply a combination of data from disparate sources into one presentation. Web 2.0 sites mash up data all the time without asking their users to do much. Problems only seem to arise when when users are required to link data together in manual, custom ways.

Intel Mash Maker sits in your browser as a toolbar that can be expanded into a sidebar as well. As you browse the web, you’ll come across sites for which it already has mashups available. For example, if you go to your friends page on Facebook, it will suggest that you activate a mashup that shows all your friends’ avatars on a Google Map. If you agree, it will plop this map right onto the webpage, and you can even decide to pin the map to the page for future visits. This is the easy part of Mash Maker, and it’s kind of nifty.

There are other places on the web where Intel likes to highlight its mashups, such as Expedia where a “leg room” mashup will show you how much estimated leg room you’d have on each search results flight. You can also pull up Yelp reviews when flipping through Craigslist.

But much of the web is like the Wild West and there are no premade mashups yet. That’s where presumably you are expected to come up with your own. And Intel has provided all the options you need to make a mashup from right within the browser extension. But honestly, the vast majority of users are going to have no clue or desire to learn how to make mashups with it. This is where the idea of an end user mashup program falls short, and it’s same the reason you’ll never hear your (normal) friends mention how they stayed up all night playing with Yahoo Pipes.

When it comes down to it, it’s probably not terribly important that Mash Maker will have limited appeal. I still can’t quite figure out why Intel even decided to produce something like this, since they seldom (never?) release pure web products that have nothing to do with hardware. But hey, why not – I’m sure at least a few people will get a kick out of it.

Ed: Intel as a software company? Or social media to join the conversation streams?

Mash-up combines widgets to create a page. The page can be personal like My Yahoo; or shared like Facebook profiles. Google Gadgets and Facebook Apps simplify gadget creation for software developers. Intel targets widget creation by common users. Could be a tough challenge.

Sprout: Now A Viral Distribution Platform for Web Services

Ed: Integrates widgets, flash, and data portability. Also, Crunchbase integration with Techcrunch increases cross links with every copied post.

Mark Hendrickson

Sprout - the impressive browser-based tool that debuted at DEMO in January and suddenly made it feasible for non-developers to create rich Flash-based widgets – has got even better with the release of v1.5.

One of the most promising things about Sprout was that it integrated third-party services. At launch, publishers could add to their widgets, among other components, polls from PollDaddy and maps from Yahoo. Since there are countless other web services that could potentially widgetize themselves, Sprout has decided to releasean SDK that effectively turns the tool into a platform.

As if to demonstrate the SDK’s possibilities, the company is debuting support forTwitter, Seesmic, and Google forms (which allow you to collect information from users in a Google spreadsheet). Expect others to follow on their own volition.

In addition to the SDK, Sprout is adding some simple animation effects for when users interact with the various parts of a widget. You can now, for example set shapes to bevel or blur when someone clicks or hovers over them. Simple undo functionality has also been added, although I’m still waiting for proper layers.

Sprout is going from strength to strength, and as word about its capabilities spreads further, we can expect it to become quite the acquisition bait. This is a marketer’s dream tool in an era of social media/advertising, and Sprout shouldn’t have any trouble monetizing its service given that widgets are effectively advertisements for many publishers. Now with the SDK, Sprout should also attract the attention of many startups, too.

The company says that 30% of its widgets have been posted to MySpace, 20% to Facebook, 25% to other social media sites, and 25% to general websites. The largest segment creating Sprout widgets are labels and artists, followed by religious and non-profit organizations, venues, and events. See a sample widget from Universal Music (which unfortunately I can’t embed because it auto-plays music).

minimize

Sprout

Website:	sproutbuilder.com
Location:	Honolulu, Hawaii, United States
Founded:	January, 2008

Sprout is an online WYSIWYG editor for Flash that as of March 2008 has been released in a public beta. Designers can use Sprout to create, publish and track Flash widgets, websites and… Learn More

Information provided by CrunchBase

NEWS: Amazing Ad Recall Rates from Podcasts, Online Video Shows

Amazing Ad Recall Rates from Podcasts, Online Video Shows

Ads in podcasts are three times as effective as “traditional” online video ads and seven times more so than television ads, according to four campaign studies conducted by Podtrac and TNS.

Compared with industry benchmarks of 21% for streaming video and 10% for television, unaided ad awareness for podcasts was an impressive 68% on average; aided recall was even higher, at 89% on average. Embedded ads in online video shows are also highly effective, the study found.

“The high unaided ad recall figures are no doubt the results of a less cluttered environment,” said Doug Keith, former vice-president of the Media & Entertainment Group at TNS and now president of Future Research Consulting.

“Audiences are paying close attention to show content and the embedded ads within them, which greatly increased ad effectiveness,” he added.

Other major findings:

• On average, fully 69% of audience members have a more favorable view of in-show advertisers.
• The study found a 73% increase in likelihood, on average, that a consumer would use/buy an advertised product after being exposed to an ad.

Those results indicate successful targeting and the strong influence of show hosts over their audience, according to Podtrac.

The studies spanned two years (Feb. ‘06 to Mar. ‘08) and multiple product categories, including television programming, automobile, financial services, and digital imaging.

Less than 4% of all internet and 1% of TV ad spend currently goes to online video ads, according to eMarketer, but the company forecasts a tripling of current ad spend to reach $4.3 billion in 2011, also predicting podcast monthly audience size to be 50 million by 2010.

“The unaided awareness level of 68% is considerably higher in podcasts and online shows than in other offline and online media,” said Mark McCrery, Podtrac’s CEO and cofounder.

“Narrowly targeted audiences, appropriate ad formats, content relevance, and show host involvement are some of the factors which come together to produce highly effective ads in online shows and podcasts.”

Ad formats studied consisted of embedded :15 and :30 host-read audio and video ads, embedded :10 and :15 produced video ads, ad banners on publisher websites, and social networking in the form of publisher blog entries about advertised brands and related topics.

About the findings: Podtrac worked with TNS to develop the methodology for the four studies, spanning two years (Feb. ‘06 to Mar. ‘08), involving 6,869 audience members and 52 online shows and podcasts, including the following: This Week in Tech, MacBreak Weekly, net@night, and Daily Giz Wiz all with TWiT.tv’s Leo Laporte; MacBreak, MacBreak Tech and This Week in Media with PixelCorp.tv’s Alex Lindsey; Geek News Central with Todd Cochran; SDR News with Andy McCaskey;Feast of Fools; Filmspotting; and FlashTV.

NEWS: A Web Shift in the Way Advertisers Seek Clicks

Ed: Buy quality views, cheap – not text clicks. Google, Yahoo, MSN, NY Times loses.

A Web Shift in the Way Advertisers Seek Clicks

By STEPHANIE CLIFFORD

Published: April 21, 2008

Tyler Townsend, a digital media manager who plans online advertising for travel clients at Ypartnership, an agency in Orlando, had $150,000 to spend on behalf of a Caribbean island’s visitors bureau. And this client did not care about branding — it wanted action.

So Mr. Townsend, who once might have made a simple buy on a site like Yahoo, created a complex campaign, which ran in March. He bought ads on Budget Travel, and he bought out Lonely Planet’s home page for a week. He used custom ad networks that included travel-themed sites, and another that would put the ads only on high-end sites.

Last year, Mr. Townsend said, many clients were happy to spend money just to raise awareness. Since January, however, “everyone’s retail-oriented. They want as many clicks for the dollar as possible,” he said.

So far, the threat of a recession has not slowed the migration of ad dollars to the Internet — as Google’s strong results showed on Thursday, when it reported a 30 percent jump in net income for its first quarter. But as Mr. Townsend’s campaign suggests, the slowing economy might be changing where those ad dollars are being spent.

Increasingly, marketers are looking to ad networks, which sell display advertising across groups of Web sites. Some networks offer targeted advertising; others, called vertical ad networks, include sites that focus on one subject, like travel or sports.

Their growth could mean a lower share of advertising for portals like AOL and particularly for Yahoo, which is particularly strong in traditional display advertising. (Yahoo will report its quarterly earnings on Tuesday.)

In 2007, United States revenue growth slowed at three of the four major portals (Yahoo, AOL and Google) according to an analysis by eMarketer. The fourth is MSN. Any downturn could also be bad news for media sites that attract a lot of display advertising, like CNN.com or nytimes.com, at premium rates.

In the United States, $21.1 billion was spent on online advertising last year, up from $16.9 billion in 2006, according to eMarketer. Search advertising — Google’s stronghold — is the majority of that spending, according to Jeffrey Lindsay, an analyst at Sanford Bernstein.

According to a report by Imran Khan, an Internet analyst at JPMorgan Chase, ad networks “are growing much faster than the general graphical advertising industry.” He estimated that the top 20 ad networks had earned $2 billion in 2007, or 14 percent of the display market.

The reasons ad networks are thriving are price and improved technology. Ad networks charge much lower cost per thousand ads served (known as CPMs), as low as $4 on an ad network with some targeting, compared with $40 and up for some ads on premium sites like MSN or Yahoo.

“While the home pages are still very effective media buys, the price tags on them have become a little outrageous for many advertisers. For all the growth that has gone on from a site standpoint, there are other ways to amass that type of audience fairly quickly that are more efficient,” said Margaret Clerkin, the chief executive of Mindshare Interaction, a media-buying firm.

The improved technology has helped. Ad networks once served ads to pages where no advertiser wanted to be, like pages that get few hits or those with controversial content. Now, though, many attractive sites are not major home pages. Also, many ad networks now offer targeting (as do portals, for a higher price), matching ads to likely buyers.

For example, if an airline wants to promote a flight from Dallas to Chicago, it can direct those ads to users with Internet addresses from those areas. Marketers can also direct ads by content: a reader on a cellphone ratings site is probably looking to buy a new phone, while another reader on the Yahoo technology home page might be browsing stories about Wi-Fi and not looking to buy anything.

David Metter, chief marketing officer of MileOne/Atlantic Automotive, a group of car dealerships with over 3,000 employees, said that it was possible to “blow your spend” on a home page ad.

“I would much rather get more specific and go to the customers and have maybe less eyeballs and higher quality eyeballs, or less leads and higher quality leads,” he said.

Based on the success of ad networks, some big players are buying their way into the game. Yahoo bought BlueLithium for $300 million last September. Last July, AOL bought Tacoda for a reported $275 million, and in November, it bought Quigo Technologies for a reported $350 million. Last May, Microsoft bought aQuantive, which owns some ad networks through a subsidiary, for $6 billion. DoubleClick, which also owns ad networks, was acquired by Google for $3.1 billion. Dozens of other networks have sprung up — one tally at eConsultant lists more than 80.

“There’s no slowdown in terms of a really targeted ad spend,” said Mitch Lowe, whose network, Jumpstart Automotive Media, handles ads for 12 automotive sites. (Jumpstart was sold to Hachette Filipacchi Media last year for $110 million.)

However, the reliance on ad networks mean that two of their major attractions — that they are simple and that they are cheaper — are diminished.

“These are the gold rush days now for ad networks,” said David Hallerman, senior analyst with eMarketer. “And that kind of counters the appeal of ad networks for advertisers’ agencies, which was to simplify the purchase of ads. And that’s why its unlikely that a great number of ad networks will survive.”

Even given the choice of ad networks, the continued flow of dollars online means that prices are rising.

Mr. Townsend says that his clients were happy with their campaign and want him to do a similar one. Except this time, they want him to hit the networks even harder.

“We thought,” Mr. Townsend said, “for this amount of dollars, we can get you this many more impressions than we did.”