Rakesh Agrawal's blog

On newspapers, Osama bin Laden and Google

Posted on August 21, 2007 by Rakesh Agrawal

The Los Angeles Times weighed in on the new Google News feature which allows sources quoted in stories to respond. After a bizarre lead — “Many publishers consider the Internet, and Google in particular, a greater threat to their livelihoods than Osama bin Laden” — the paper goes on to slam the practice concluding that Google’s efforts are “not journalism.”

It may not be journalism in the traditional sense, but allowing sources to comment on stories that they were part of can help further the quest for truth, which is the purpose of journalism.

Journalists and news organizations have for a long time fought to prevent the transparency and accountability into their work that they demand of government and other corporations.

Instead of fighting such attempts, the media ought to embrace transparency. They should link to original sources whenever possible. This includes press releases, audio from phone or in-person interviews, government documents, etc. For too many stories, it will reveal that the “reporter” has done little more than move paragraphs around from the press release. For other stories, a reader will be able to determine that a quote was taken out of context or the meaning significantly altered. Good. These kind of practice needs to be exposed.

And there will be plenty of stories where the additional content adds credibility to the story and provides more depth for readers who are passionate about the topic.

This kind of transparency will help to increase the credibility of good media organizations and damage the credibility of those who play fast and loose with the facts or take dangerous short cuts.

The Times worries that the comments section in Google News “is likely to be larded with spin, hype and obfuscation” and “won’t help readers separate the factual wheat from the public-relations chaff”. As long as sources are clearly identified, I trust the readers to make that distinction on their own.

As the gatekeepers of what goes into print or on the air, news organizations have historically had tremendous power over the public’s perception of events. Regardless of what Google News does, the Internet has dramatically changed that by giving sources and the public the opportunity to respond. Dan Rather found out the hard way when bloggers exposed serious concerns about the authenticity of documents used as the basis of a report on Bush’s service records.

In the tech world, reporters who misquote a source can pretty much count on having the source call them on it in a post on the source’s blog (or in a comment on the story itself). It’s only a matter of time before other industries catch on.

Recommended reading:

The L.A. Times tells its readers: ‘Shut up’ (Online Journalism Review)
Don’t like how a reporter treated you? Tell Google (previous coverage from me)
Perspectives about the news from people in the news (Google News blog)
Why Google’s “News Comments” Idea Will Fail (PR Squared blog)

Posted in google, journalism, media, newspapers | 1 Comment

Revolutionizing journalism education at Medill

Posted on August 20, 2007 by Rakesh Agrawal

Chicago magazine has a piece on the challenges facing John Lavine, the new dean at Northwestern’s Medill School of Journalism as he sets a new course for teaching journalism.

“It would be unethical for us to educate you to only be able to write,” [Lavine] said. “It would be like sending you out with your left arm and your right leg tied behind your back.”

After taking over as the leader of Medill earlier that year, the new dean had vowed to “blow up” the old curriculum at what has long been considered one of the best journalism schools in the country. He declared that students needed to be immersed in “new media”—Web sites, videos, filmstrips, video games, and podcasts. And the new curriculum would emphasize an understanding of “audience”—who the customers are, what they want, how to reach them. The concept of marketing—widely disdained by ink-in-their-veins journalists—would assume a key role in the teaching program.

I graduated from Medill in 1995. I was fortunate to take a class called Newspapers of Today & Tomorrow. That class placed a heavy emphasis on prototyping and knowing the audience. It also gave me the freedom to explore online newspapers, leading me to launch The Daily Northwestern online.

Given today’s $35,000 annual tuition and the challenges facing the media business, I’d be hard pressed to encourage someone to go into a program that only taught writing and editing. As it is, I’m constantly encouraging my friends in the media business to learn about their audiences, engage with their readers and learn new ways of telling stories.

If no one wants to read your work, it doesn’t matter that you know “who” from “whom.”

Posted in journalism, media, newspapers | Comments Off

comScore redefines search, Google wins bigger

Posted on August 20, 2007 by Rakesh Agrawal

ComScore is changing the methodology for its qSearch market share ratings. Instead of just counting search activity at the major search engines, comScore is expanding the definition of search to include searches at sites such as Wikipedia, eBay, Amazon, MySpace, Mapquest, Craigslist and other vertical players.

Searches across multiple tabs for the same search term will also be counted separately. For example, if you search for “hurrican dean” in Web search and then click the tabs for news and pictures, that will be counted as three searches.

For those who were hoping this might shrink Google’s share of search, think again. Under the new methodology, Google’s share grew 6 points in March compared with the old methodology. The additions to Google (which include YouTube) are greater than all of TimeWarner’s search traffic (which itself benefits greatly from the addition of Mapquest).

Here is a comparison of core search and expanded search metrics based on July 2007 data:

Core search	Expanded search
Google Yahoo! Microsoft Ask Time Warner (AOL Search)	Google (Google, YouTube) Yahoo! Microsoft Time Warner (AOL Search, Mapquest) Fox Interactive (MySpace) eBay Ask Craigslist Amazon Infospace

Using the expanded definition, Ask drops from #4 to #7, being passed by TimeWarner, Fox Interactive Media (MySpace) and eBay. TimeWarner moves up from #5 to #4, based largely on Mapquest traffic.

The numbers don’t seem to include Facebook, which according to its blog does more than 600 million searches a month. If that number were comparable to qSearch data, Facebook would be at #5 in the expanded search.

More on: AOL, Google, Yahoo!, Facebook.

Disclosure: I used to work at AOL Search.

Posted in aol, facebook, google, metrics, search, statistics, yahoo | 3 Comments

Rewriting your corporate history on Wikipedia

Posted on August 14, 2007 by Rakesh Agrawal

WIRED has a story about companies rewriting their history on Wikipedia. The story is based on Wikipedia Scanner, a tool from Cal Tech grad student Virgil Griffith that allows you to look up anonymous Wikipedia edits from a specific company’s computers. Among the companies who apparently edited their own Wikipedia entries are Diebold (removing criticism of its voting machines), Wal-Mart (outsourcing, employee wages, etc.), Exxon (that Valdez thing) and Microsoft.

Some of this appears to be transparently self-interested, either adding positive, press release-like material to entries, or deleting whole swaths of critical material.

Voting-machine company Diebold provides a good example of the latter, with someone at the company’s IP address apparently deleting long paragraphs detailing the security industry’s concerns over the integrity of their voting machines, and information about the company’s CEO’s fund-raising for President Bush.

The text, deleted in November 2005, was quickly restored by another Wikipedia contributor, who advised the anonymous editor, “Please stop removing content from Wikipedia. It is considered vandalism.”

It’s not just corporations; religious groups and politicians are also cleaning up their own images. See WIRED’s Threat Level to view and vote on the most shameful spin jobs.

What’s amazing is that they’re not even trying to hide it; the changes are being made from trackable locations. If the same changes were made from a coffee shop, they wouldn’t be directly trackable.

There’s no direct evidence of intent. It could be corporate policy to edit Wikipedia for public relations or it could be an employee who takes the initiative without company approval. Or it could just be someone goofing off on company time: “One CIA entry deals with the details of lyrics sung in a Buffy the Vampire Slayer episode.”

Posted in web 2, web 2.0, wikipedia | Comments Off

AOL launches improved mobile search

Posted on August 14, 2007 by Rakesh Agrawal

AOL today released its new beta of mobile search. Congratulations to rockstar developer Alan Tai and product manager Farhan Memon. Alan did much of the initial prototyping on his own time while we worked to get approval.

I pushed the strategy on this, so it would be inappropriate for me to review it. See Om’s blog for more details.

I’ve long believed that you need to design for the medium. Shovelware didn’t work when we were first trying to put content on the Web; it won’t work now. The old version of AOL’s mobile search took the same 10 Web results you would get on a Web browser and shrunk them down to fit a mobile screen. That didn’t work.

People are in a different state when they’re mobile. Most people aren’t going to do research for a term paper or browse real estate listings on their cell phones. (Not least because most of the sites won’t work well when shrunk down to fit a mobile device.) The new mobile search is designed around answering the questions that people are most likely asking when they’re out and about: What’s the weather like? What’s the phone number for the local pizza place? What time is the movie starting?

Then there are the issues of limited screen space and difficulty in entering data. Time to answer is especially critical in mobile. This product was designed to get people answers to common mobile queries as quickly as possible.

Posted in aol, local search, wireless, wireless data | Comments Off

Google’s $4.55 bag of cookies

Posted on August 11, 2007 by Rakesh Agrawal

Vending machine priced by grams of fat, Google, San Jose, California.jpg, originally uploaded by gruntzooki.

I was visiting my friend Adam at Google yesterday and he pointed out a vending machine in Google’s Building 43. A vending machine on the Google campus? Isn’t all of the food free? Most of it is. (Thanks for the free lunch, Adam!)

This machine is an exception. And unlike most vending machines, the food is priced based on how bad it is for you. Items high in fat and sugar cost more. The most expensive item I saw was a $4.55 bag of Famous Amos cookies. (You can see it at the right end of the second row.) It’s a terrific way to illustrate the externalities that we don’t take into account when buying food.

I would love to see schools use something similar to illustrate healthful eating habits to school kids. That could be the compromise position between parents who want to ban vending machines and vending machine companies that want nutrition education. (It will never happen.)

Food marketing is based on increasing profits and not health. At restaurants, convenience stores and similar places, they do whatever they can to get you to consume more because it adds straight to their bottom line while it adds straight to your waistline.

I once heard a professor on a radio show asking if an 8″ pizza costs $10, how much should a 16″ pizza cost? Most people would answer $20, because they (incorrectly) thought that the 16″ pizza is twice the size of the 8″. The “correct” answer to the question was $40, because a 16″ pizza actually has 4x the area of an 8″ pizza.

He’s clearly a math professor. The marketing and profit maximization answer is somewhere around $14. Most of the cost of the pizza is in telling you that they sell pizza, the fixed costs of operating the store and delivering it to you. The incremental cost of a 16″ pizza versus an 8″ pizza is negligible. The bulk of the extra $4 is pure profit. The goal is to set the price difference low enough that you feel like a chump if you buy the 8″ pizza.

Marketers also use naming to influence your consumption. Five Guys, a hamburger chain on the East Coast, offers a choice of a “Cheeseburger” or a “Little Cheeseburger”. What guy is going to order the “Little Cheeseburger”? On the other coast, In-N-Out Burger doesn’t provide a financial incentive to consume more. Their combos cost the same as the individual components put together.

More on: Google

Posted in advertising, fun, google, random | 6 Comments

Why we love social networks

Posted on August 11, 2007 by Rakesh Agrawal

There was a comment in Paul Kedrosky’s blog last month asking about the real value of social networks:

Facebook, MySpace, I don’t get it: personalize a webpage template and exchange links, I mean friendships, with people.

Since I responded to the comment, I thought it was worth expanding on and posting here. Here are some of the key reasons I believe people have taken to services like Facebook.

People like to talk about themselves. (Including me.) The explosion in blogs is just one example. But blogging is a lot of work. You have to pick a blog platform, pick a name for the blog and write posts. Each post is a fair amount of work to assemble. (I spend 20-45 minutes on a post, depending on how much research is involved.)

At AOL, I talked frequently about microblogging — allowing people to quickly and easily express their thoughts without all the overhead of blogging. Lowering the bar increases participation.

That’s exactly what social networking sites do. I can go into the Flixster Movies app on Facebook, search for a movie, click a star rating and write out a one-paragraph review. The actors, title, summary and movie picture are pre-populated. By contrast, every time I come to WordPress I face a blank screen. It’s like the difference between a fill-in-the blank test and an essay question.

People want to feel connected, but they don’t want to do a lot of work to be connected. Social networks allow us to keep in touch with many more people than we could ordinarily keep track of. I have many former colleagues, college friends and other acquaintances who I want to stay in touch with. I want to know what’s going on in their lives, but I don’t have time to call and email everyone. One of my favorite applications is the slickr screensaver; it taps into flickr and lets me see the latest in my friends lives when my computer is idle.

These networks also help me discover common interests with people I know. When I visited Carl Kasell’s profile on Facebook, I found that six of my friends are also Carl fans.

There’s also the value of persistence: with changing email addresses, phone numbers, jobs, it’s not always easy to find someone. On LinkedIn and Facebook, I’m connected to the person, not the specific contact address.

Email today is a disaster. With all the spams, scams and other nasties, the closed environment on Facebook is a godsend. I know when I get a message on Facebook that it’s most likely not spam. There’s no chance that the message will get trapped by an overly aggressive spam filter (I get way too many false positives) — you’re more likely to reach me through a comment on my blog or a Facebook message than by email.

Discussions elsewhere have gotten out of hand. Pretty much any blog post or discussion that gets more than a few commenters on most sites devolves into personal attacks. When conversations involve real identity the discussion is usually (though certainly not always) more civil.

As the song goes, sometimes you want to go where everybody knows your name, and they’re always glad you came.

More on: social networking, Facebook

Posted in email, facebook, social networking, web 2, web 2.0 | 1 Comment

Virgin America takes to the skies; Chron doesn’t

Posted on August 10, 2007 by Rakesh Agrawal

Wednesday marked the inaugural flights of Virgin America, a new low-cost airline based in San Francisco. Virgin America is currently flying from San Francisco to Los Angeles and New York. This fall it will add flights to Las Vegas and Washington Dulles.

From all accounts, Virgin America is setting a new standard in airline amenities. (See some of my earlier coverage on Virgin America.) Many of Virgin’s amenities are especially appealing to geeks. The geek bible, Engadget, provides detailed coverage of the maiden flight from JFK to SFO. Their coverage includes 136 pictures.

WIRED also provides detailed coverage with additional photos.

Engadget and WIRED both had reporters on the inaugural flights. Engadget flew from New York and paid for the flight; WIRED flew from Los Angeles and took a freebie.

Based on the coverage in the San Francisco Chronicle, it doesn’t look like it had a reporter on either flight. It ran a piece with two staff photos and a Reuters photo from New York and an AP photo from SFO (!).

This is a huge business story for San Francisco and the Bay Area. The airline has added 500 employees (most based in the Bay Area) and expects to grow to 5,000. Yet the Chron got beat out by a gadget blog and a WIRED blog. Heck, I even considered taking the LAX flight and paying the whopping $44 out of my own pocket. (I couldn’t get the schedule to work out.)

I was at a panel discussion Thursday night sponsored by the Social Media Club featuring Kevin Rose of Digg, Evan Hansen of WIRED and Chris Tolles of Topix. The audience consisted of many in the old media who expressed worry about new-media types stealing their content. If they’re going to get beat like this, they won’t have to worry about that for too long.

Posted in airlines, journalism, media, newspapers, travel | Comments Off

Don’t like how a reporter treated you? Tell Google

Posted on August 9, 2007 by Rakesh Agrawal

The Wall Street Journal reports that Google is allowing newsmakers to comment on stories that appear in Google News.

Google says it allows people who are quoted in articles or affiliated with organizations in them to comment, as well as reporters and editors responsible for the stories.

Google will be authenticating the identities of those responding.

For this story about kids preferring food in McDonald’s packaging over identical unbranded food, there is a comment from a flack from McDonald’s and a professor of pediatrics. (You can also read my analysis of the McDonald’s packaging story.) I think it adds a valuable dimension that the reporting of the original stories lacked.

It’s another step in the path to making news an ongoing conversation instead of just a one-shot push to the reader/viewer.

I haven’t had a chance to dig into it yet, but I definitely will.

More on: journalism, newspapers

Posted in google, journalism, media, newspapers | 3 Comments

Writing news for search engines and blogs

Posted on August 9, 2007 by Rakesh Agrawal

One of the reasons I love blogging is that it gives me the opportunity to see things at a micro level. I can see patterns and analyze data in a way that I couldn’t in a typical work role.

When I was writing the follow on post to the 35W bridge collapse the other day, I initially wrote this:

MN-DOT has finally released the video from last week’s bridge collapse.

That’s how I would have written it based on my journalism experience. As I tapped out the period, I realized that the sentence is meaningless to search engines. And thus unfindable by the many users who rely on search engines to find news. I rewrote it as:

MN-DOT has finally released the video from the August 1 collapse of the Interstate 35W bridge over the Mississippi River.

That sounded too stilted to me. Based on having looked at my traffic data, I knew people weren’t searching on “August 1”. They were searching heavily on “Mississippi” and “35W”. The final version I used is this:

MN-DOT has finally released the video from last week’s collapse of the Interstate 35W bridge over the Mississippi River.

Writing headlines for blogs is even trickier. Blog headlines have two audiences: search engines and readers who view blog posts in RSS feeds. The clever headline that might get a reader to click on a link is often lacking in the keywords that search engines need.

This headline from Tuesday is meaningless to search engines: “Mmmm… McCarrots and McMilk.” It seems to be working from an RSS feed clickthrough perspective.

I strive for a mix of people friendly and search engine friendly headlines. When I use headlines like the McCarrots and McMilk one, I do an extra pass to make sure that the body of the post contains the keywords searchers are looking for.

Speaking of search engines and news, this week drove home a pet peeve: news sites like CNN defaulting their search engines to search the Web. If I wanted to search the Web, I would have gone to Google, used the Google search box in Firefox or used the search box in the Google Toolbar. If I’m searching on your news site, I want your news content.

I got some traffic to this blog from people searching CNN for “35W bridge traffic camera video”. That search led them to this post on the video being released. Which led them back to the video on CNN’s site.

As much as I welcome the traffic, it’s a terrible user experience. A hybrid model, where CNN content comes before Web results, would be more effective and still serve the revenue goals of offering Web search.

More on: newspapers, journalism