Archive for the ‘Google’ category

Interact08: Marissa Mayer, Live!

29, September, 2008

Google’s Out-in-Front Woman, speaking at Interact08. High points via liveblog:

1. 20 percent time. At one point, 50 percent of Google’s newer products came from the “20 percent time”–the 20 percent of time all Google staffers are allotted to work on personal stuff with no relationship to their current “real” work. Drives constant innovation.

2. Beyond the wiki: Internal 2.0 at Google HQ: Every Monday, people file their “snippets”–5 bullets of what they’re working on. All are submitted, published, searchable–transparent organization to share knowledge easily.

3. Behind the curtain. Shows a diagram that illustrates how a query moves from end user, through the massive Google backend, and back to the user. How? Via 400-1,000 machines! All in .2 seconds! Um, cool, in a hyperspeed 1999 way.

4. Castle-building vs. iterating. Instead of building the One Great Thing for years [as Apple famously does], Google launches products  “early and often,” driven by response to users.

5. How do you iterate? Use data to drive decisions. “Data is apolitical.”

6. So: Split A/B testing [n.b. a case of spontaneous generation; Amazon came up with the same technique at the same time]. Different users see different products or results, and the outcomes of these interactions drive future iterations. Great example about how subtle differences in vertical spacing had effects on search behavior. Also, via same technique: users liked blue, not yellow, boxes for search ads. Also, Google launched with 10 results per page. When they put more results on page to see if that’s the right number, searches dropped 20 percent. Why? Slower delivery of results, as tiny as those differences are, discourage people. And: Adwords succeeded because of immediate posting of ads.

7. Future of Search? One Saturday, Mayer kept track of all her searches that didn’t produce results. She showed a whole presentation screen to show all the failures. Among them: What is the largest city in Russia after Moscow and St. Petersburg? “J.C. Penney” near an obscure town in South Dakota, many dozen others.

So how will search improve over the coming years?

New Modes

Ubiquitous mobile–in car, wearable devices, eyeglass displays, watch that keeps track as you speak and produces results. You can embed search into everyday life.

New inputs–why can’t you talk to a search engine, or type in natural language? Why can’t you use an image or tune as your search query?

New Media

Google search results currently return 10 urls. Maybe it can produce 10 “answers”? Or: Now we return some video for proper queries [“how to do the charleston dance” query produces a video that shows you how]. “How to tie a bow tie” is better–the search currently returns diagrams and demos. Media needs to respond more to the nature of the query.

Personalization

The coming engines will understand more about the user–based on geography, past searches, personal preferences, who your friends are, other contextual information, etc.

Takeaway impressions: Google’s process of persistent iteration explains why they remain market leaders. The use of “20 percent” time drives the company forward. Mayer is brilliant.

Damn. I hate it when Google looks indestructible.

Knol and SEO: Google Insider Trading?

11, August, 2008

Is Google a media company? The New York Times asks the question today. The report focuses on Knol, Google’s new super-publishing platform which invites people to post articles online.

Authors retain copyright, and Google makes no money directly for the postings, though authors can choose to display AdSense ads on their content pages. Google serves those ads and shares in their revenue. This is not an insignificant part of the Knol strategy.

By asking whether Google will artificially elevate Knol content on its search engine results pages [SERPs], the Times is, I believe, asking the wrong question.

Google doesn’t need to employ illegally anti-competitive means. Knol content likely has the advantage of perfectly legal, perfectly legit, perfect SEO. Because who could possibly know SEO better than Google programmers and engineers?

For the uninitiated: Search Engine Optimization is the art/science of enhancing content [and the sites that carry it] to make that content appear higher on search results pages. Some people are SEO wizards, whose long experience, wily intelligence, subtle understandings of code and ability to read Brother Google’s head-fakes give them the ability to make just about any content, including total crap, pop up really high on SERPs. We mortals follow half-a-dozen best practices and muddle through. [Here’s an independently authored Knol [!] on how Google’s page rank system appears to work. Here’s another Knol on SEO generally.]

But here’s the problem: Google keeps its algorithm secret, so people on the outside never really know for sure what Brother Google’s search bots are sniffing for. That’s why SEO experts exist. If Google’s search algorithms were transparent, we’d all know the deal. SEO experts would be neither necessary nor well-paid.

From the Times piece:

Google has always said it will never compromise the objectivity of its search results. And it says it treats Knol pages like any other pages on the Web. “When you see Knol pages rank high, they are there because they have earned their position,” said Gabriel Stricker, a spokesman for Google.

Yes, but Google doesn’t have to compromise the objectivity of its search results to make Knol content rocket to the top.

If we generously accept the claim that the Knol pages have “earned” their way high onto the SERPs–that Google has, in the name of fairness and despite powerful economic motivations, heroically resisted the temptation to tap its internal knowledge of how Google’s algorithms reward and punish web pages when building its platform–then we are entitled to know how the Knol pages have done so well via conventional SEO.

But the Mountain View mothership is eerily silent about how it is indexing and making Knol content discoverable. Here’s an answer from an author of Knol’s Help pages to a user question:

“Q: . . .What factors affect whether a page appears [on SERPs] or not?

I cannot answer these questions, only to say that we are dedicated to making the search experiences for readers and authors better, over time.”

As discussions among SEO folk are revealing, some Knol entries are showing up inexplicably high on SERPs already. The Times story reports that a buttermilk pancake recipe published as a Knol is appearing higher on SERPs than a Martha Stewart Living page that publishes a similar recipe.

If this is not due to some internal knowledge of Google’s algorithms, the only explanation could be super-duper-ultraly-awesome-galactic-class search engine optimization practices by Google based on publicly available information which the greatest minds in SEO otherwise have not discovered.

For instance: SEO experts coach that links to a site’s content from other credible web pages are the most important factor driving SERP rankings. Yet some Knol entries that have been up for two or three weeks and appear to have just a few inlinks are scoring higher than similar Wikipedia entries–or content from well-established publishers whose content is richly inlinked and appears on pages with what’s known in the trade as “high URL equity.”

Yet Google won’t explain this oddity other than to say it’s not cheating. It won’t comment any more than that.

So. Let’s think for a minute about how Knol is pulling this off if it’s not using inside knowledge. . . .Ah: Maybe Google has hired brilliant SEO experts from the outside to make Knol entries rise high on SERPs so quickly. After all, that’s what the rest of us have to do.

So, Brother Google: If you won’t produce an explanation, at least produce an invoice that shows you’ve hired that outside SEO firm. I have to say, whoever is doing SEO on Knol content is doing a damn good job.


Bookmark and Share

Viewzi’s Visual Search: I’ll Know It When I See It

22, June, 2008

Let me be unambiguous: It’s Google’s world, we just live in it. There is no “search war,” no “game-changers” in the world of search. When the End of Days finally arrives, some bony finger will type “eschatology” in the search box, hit “I’m Feeling Lucky,” and the world will end. Google’s victory will be complete.

Happily, none of this is preventing people from doing some wily, aspirational things with search. The most compelling (if maddeningly flawed) example I’ve seen is called Viewzi, which has just opened itself to the public after a buzzy closed beta.

Short version: It’s a visual search tool that offers 15 [!] different ways to view search results. It’s a dazzler, a hum-dinger, a Halloween bagful of eye candy. If you’re a flash developer, a dataviz geek or a distractable noodler, you’ll find it irresistible. Viewzi makes Google’s results look like Braille.

Put a query in the search box, and a ribbon of blurry choices spreads across the screen: Basic Photo View, VideoX3 View, 4 Sources View, and more. [Note: Since this is an application built in flash, I can’t provide specific URLs to any of these features. If you click on the images below they’ll take you to a new search box. You’ll need to conduct a search yourself to see the features I’m discussing.]

Viewzi Mix

Below is the 4 Sources view, which presents screen shots of results harvested from Google, Yahoo, Live and Ask. I can’t understate the goofy pleasure I get rearranging and digging among these results. Bonus: You can see immediately which results the engines share, value differently, bury, etc. SEOers will dig it.

Viewzi 4 Sources View

But the most powerful–and potentially disruptive–feature is something called 3-D Photo Cloud view. It has a creepy, responsive intelligence that I find affecting in ways I can’t explain. It somehow creates the unsettling impression of knowledge accumulating in real time, of neural pathways proliferating as you watch, of an infobeing gathering power as it grows. [I have not been drinking anything stronger than coffee while writing this, I swear. This thing is freaky.]

Yakov Sverdlov, Viewzi 3-D


The Viewzi project has the feel of an open-source playground, a platform where search geeks and datavizualists can create new ways of organizing information visually. This may turn out to be the real value of Viewzi–a kind of Challenge X for visual search that inspires some serious bug-eyed innovation. [Or not: There’s already evidence of creativity being stretched thin over commercial ambitions: There are Celebrity Photo, Weather, Recipe, Shopping and TechCrunch (?) views. Can a FaceBookNewsFeedView (sm) be far away?]

Meantime, I tried Viewzi for some “real” searches I’d recently done on health, a recent political poll, an old friend from college, some tax stuff, a vintage car. Here’s what I realized: Most searchers are harshly pragmatic, unforgiving of excessive keystrokes and distractions. Google is perfect for the drive-by infosnag.

Viewzi offers some simple search views for mundane topics, the most servicable of which is the Web Screenshot View, which allows you to scroll through images of results pages. It’s slower and more annoying than Google, but it allows you to preview a source before you click into it.

So That\'s a Matador?

Google rules the everyday search. But if you have the need or leisure to dig into a topic and explore it from a bunch of different sides, Viewzi has plenty to offer. Block out two hours on Outlook and close your door. You’ll be awhile.

But if anything funny crawls out of that 3-D  Photo Cloud and attaches itself to your forehead like a tick, don’t blame me. I warned you.

A Google Health Clinical Exam

22, May, 2008

Not one more pixel need be spilt about the issues of privacy, security, HIPAA, metastatic data, third-party crashers, or corporate imperial overreach raised by the debut of Google Health. Let’s just snap on the latex gloves and do a quick exam. This won’t hurt a bit.

Three brief clinical observations follow:

Your conditions, your choice

You can enter your “conditions” either by entering text or choosing from a disheartening alphabetic menu of bodily afflictions, from Aarskog Syndrome to Zollinger-Ellison Syndrome. The list is 20 screens by 3 columns deep when spread out on one endless page.

Immediately preceding the last entry is “Zits”–a nice bit of diction that helps reach users where they live, so to speak, to humanize the Google Machine. As with many conditions that populate the picklist (no pun), there’s a pre-loaded search for zits. But only certain conditions are pre-loaded with searches. Although “whiteheads” was on the list, when I typed it in there was no stored search. When I did the search myself up popped the zits search results.

To give the product a test run as you can see below I chose a number of conditions from the list — WHICH, IF YOU ARE AN INSURER, EMPLOYER OR ACQUAINTANCE, I ASSURE YOU ARE ENTIRELY MADE UP AND DON’T APPLY TO ME AT ALL, IN FACT I AM PERFECTLY HEALTHY. I also tried to throw Brother Google a curve ball by describing the same conditions using several different terms, i.e., arthritis, osteoarthritis and bad knees. I was permitted to add these as I wished. To see if there was any filtering or databasing of my conditions going on, I outright invented a disease by lashing together some of my favorite Greek roots: pyohemoflatalgia. (Go ahead, look ‘em up.) Brother Google didn’t blink.

\

Conclusion: Google isn’t databasing my conditions. I’m just entering text, and the alpha picklist is just there to prompt people to identify diseases by common names.

Google Health “Research” Offers Less–and Therefore Better–Content than a Google Search

Hit the “research” link accompanying any condition and up pops a neatly tailored page on the topic. Most of the page (on, say, osteoarthritis) consists of a spectacularly workmanlike article from the utterly competent information provider A.D.A.M. The right rail has a set of links curated by some unknown hand or machine. Depending on the topic you may get a few blurbs of news items (by some method culled from the longer, messier, far less coherent Google News results on the same topic), links to Google Groups, Google Scholar articles or related searches (“search trends”).

Runner\'s Knee Sidebar


Observation: The regular–i.e., non-Google Health–Google searches on the same topics provide better results than Google searches on non-medical topics. (Google has for some time used a service called “Google Co-op” to serve up results from only selected health content providers.) This is good–a tacit acknowledgment by Brother Google that searches for lymphoma are more important than those for, say, “David winner Idol”. STILL, THE FIRST LISTING ON THE TOP OF GOOGLE SEARCH RESULTS FOR MANY HEALTH-RELATED “SEARCH TREND” TOPICS IS OFTEN FROM WIKIPEDIA. WOULD YOU GUYS STOP THAT?

Did somebody mention “beta”?

We all know how Google overuses the term “beta” to (correctly, often) imply an evolving product and to (necessarily, often) seek forgiveness for bugs that haven’t been scrubbed out yet.

Well: After creating “my” “personal” health “record,” I tried the Google Health (beta) find-a-doctor function. “diabetes” doctors in “Bethesda, Maryland.” Enter.

This keystroke should have triggered a klaxon audio file that screamed BETABETABETA!!! Ah-OOO-Ga! Ah-OOO-Ga!

I don’t know, maybe that top endocrinologist is staying at the Grand Hyatt Washington. Or maybe he’s taken a job with the Department of Health and Human Services? (Or is it the Department of Transportation?) Who can tell?

Anyhow, the seven minutes I have for basic clinical exams is over. It was fun giving Google Health a quick once-over. But it’s a good thing the medical record is a fake. I don’t think it’s ready for circulation yet.

In fact, there were enough suspicious observations in my quick exam that. . .I think I may have to schedule a biopsy next. Who knows what’ll turn up?

Google vs. Microsoft at the Health 2.0 Spring Fling

11, March, 2008

Here’s a reprint (with modifications) of a story of mine that appeared in The Washington Post today. Based on my recent visit to the Health 2.0 “Spring Fling” conference in San Diego, hosted by the brilliant and beautiful Matthew Holt and Indu Subayia, it runsdown what appears to be shaping up between Google and Microsoft in the consumer-PHR-platform space.

I’ll be adding other entries about Health 2.0 products in the coming days and weeks.

Microsoft HealthVault vs. Google Health

By Craig Stoltz, Special to The Washington Post

Personal health records, or PHRs, were the buzz at last week’s Health 2.0 Spring Fling conference in San Diego — especially recent entries by Google and Microsoft that have the rest of the industry energized, focused and at least a little bit frightened.

Bill Reid, director of Microsoft’s HealthVault program, described the effort to integrate information technology into personal health care as a “long journey. We’re just at the front end of the process.” Was this an acknowledgment of the complexity of the task ahead? Or a a shrewd way to reduce expectations about the software giant’s big investment?
Based on the tenor of the Health 2.0 conference–a high-energy gathering of great minds, big ideas and entrepreneurial hustlers–it may be both. In addition to Google and Microsoft, dozens of companies presented online products designed to make U.S. health care smarter, stronger and better looking. There was a plan by a firm called Organized Wisdom to offer online doctor consults at $1.99 per minute, a provider search tool pitched as “the match.com of health care,” and an electronic medical record by a firm called Myca that made you want to bask in the sheer beauty of ear infection data.
Here’s a look at where Microsoft’s and Google’s personal health record programs are now and where they may be headed.
What Microsoft HealthVault and Google Health have in common:

* Both companies claim the same ultimate goals: To create integrated online environments where you can create and store your personal records, get information, find doctors, make medical appointments, communicate online, manage medications, share information with providers and more. Oh, and with Microsoft and Google, there’s always that other goal: to dominate the world.

* Both put users in control over what goes into the record and who has access to it. If there’s something you’d rather not share with your employer, insurance company or anyone else, leave it out.

* Both are free Web-based services, meaning you can access the records without cost from any computer. The services are described as being as secure as online banking. Both companies pledge not to share your information without your explicit permission.

* Both offer tailored searches that promise to filter out garbage and surface the gold.

Microsoft HealthVault:

Debut: October 2007.

The story so far: Microsoft has assembled a list of companies that make products (glucometers, blood pressure monitors) or offer services (software that pulls in data from labs, hospitals, etc.) compatible with HealthVault. Use one of them, and data from the lab or your home blood pressure cuff automatically gets sucked into your HealthVault PHR. If you aren’t using one of these products or services, though, the only way to create your record now is by uploading existing documents — a recent page of bloodwork results, say — from your computer.

Follow the money: Microsoft plans to make money by placing ads next to HealthVault search results. As with any search, some are text ads generated by keywords. Some are interactive ads promoting HeathVault-compatible devices or services. Some offer related books and products from Amazon. Anyone can use the HealthVault search, but if you want to save your results privately (a nice feature), you’ll need to sign up for a free HealthVault account.

Curious observation: HealthVault’s search results are sometimes riddled with information from interested sources (supplement makers!?!) and below-gold-standard publishers. Do I really need a tailored, secure search to find a Wikipedia article on arthritis?

Google Health

Unveiled: February 2008.

The story so far: The first live-action test of Google’s PHR is a pilot project with Cleveland Clinic launched last month and expected to run six to eight weeks. Screen shots of the service suggest people can create their own PHRs via simple forms with check boxes and pull-down menus. Like Microsoft, Google plans to offer the ability to automatically pull in data (for example, X-rays and readouts from a pedometer) from devices, services and health-care providers. Google is encouraging use of open technology standards that will let the health world’s many different information systems talk to each other easily.

Follow the money: Google doesn’t rule out the possibility of selling ads alongside search results or other Google Health services but says it has no current plans to do so.

Next steps: After the Cleveland Clinic pilot, Google says it will digest what it has learned and move toward launch. No date is set.

Curious observation: So why would Google take on such a big, difficult project — creating complex data exchange systems and storing all that personal information — if there’s no way to make money? Data show more than 70 percent of people seeking health-care information turn first to Google. A strong personal health dashboard linked to other Google services, including its cash-cow search business, can make sure those health-seekers stay with Google rather than with the competition. Like Microsoft, for instance.

Google’s Ad Scan: Stupidity, Madness or Mere Insanity?

29, January, 2008

The blog Silicon Valley Insider has an extraordinary item today about Google’s latest ad scheme, which works this way:

1. Advertiser buys ad in newspaper; a Google barcode appears on the ad

2. For reasons unrevealed and hard to imagine, a newspaper reader wants so badly to have the ad on her mobile phone that she takes a snapshot of the barcode with her device (she has previously installed special software to permit this)

3. Her mobile phone takes her directly to an online version of the ad–which presumably has more value in mobile, digital form than it does in its mobile, analog print version. Maybe it’s more information, or a chance to sign up for, oh, I don’t know, an e-mail newsletter or a sweepstakes (which is to say an e-mail newsletter)

4. The newspaper advertiser can now capture data about that potential customer which is uncaptureable in a print environment

5. Google sends the cell phone user an e-mail message, which she is invited to print out and attach to her forehead with duct tape. The e-mail says “I Am Verry Stupid.”

Actually, I made up No. 5.

I know, I know, they do this printed-bar-code-and-cell-phone thing in Japan. But in Japan they also watch those game shows where people hilariously incur significant internal organ damage by competing in contests involving mud, spinning platforms and immovable objects. “I Am Verry Stupid” indeed.

Business Rule No. 1: If you want to get consumers to do something that benefits you (like give you money, or merely identify yourself as a reader of an ad so someone else gives you money), you have to offer them something of value in return. This “value proposition” seems to be missing in this Google scheme.

Please, right now, close your eyes. Try to imagine seeing an ad in the newspaper so utterly compelling that you’re willing to stop whatever you’re doing, take out your cell phone, and view another version of that ad–maybe an interactive one, just like on the web!–in the tiny screen of your cell phone.

I know, I can’t either.

I prefer to view a money scheme this spectacularly wrong-headed as another hopeful sign that Google has entered its Late Empire phase, when the emperors drink arsenic at orgies while the lean and crafty barbarians consolidate control of the provinces and plan the ultimate pillage.

Or maybe (alas) it’s just another sign that Google has so much money now it can make big bets on slim chances and put nothing more at risk than a quarterly earnings rounding error.

In any case, if you know anyone who works on this particular project at Google, please send them the following message as an e-mail and ask them to tape it to their foreheads.

I AM VERRY STUPID

Yahoo Beating Google! Sort of

20, December, 2007

The latest traffic figures from comScore shows that Yahoo has a comfortable lead over Google.

In a way.  If you look at the numbers in a certain light.

In the measure of unique pages views of site networks, all Yahoo sites delivered more views (136 million) than all Google sites (131 million).

In a measure of how many web users had viewed an ad appeared on on any Yahoo site, Yahoo scored 85 percent, compared to Google’s 76.

Of course, these numbers tell, at best, a part of the story. Comparing revenue, searches conducted, total reach, etc., Google has a commanding lead.

And in first place among ad networks whose ads were viewed by the most Internet viewers? Why, advertising.com,  now owned by Time-Warner’s AOL.


Follow

Get every new post delivered to your Inbox.