Entrepreneur Watch

How Google determines relevancy of TV commercials

I've long wondered how Google plans to make money off non-internet advertising. After all, how do you determine that a TV commercial, for example, is relevant to the people watching a particular program?

It turns out that Google is trying to measure such relevancy. In a blog post, Dan Sigmond, Technical Lead for Google TV Ads, explains that Google has a relationship with Echostar to track when people switch channels on their set-top boxes.

They found that "between 5-15% of viewers tune away from a channel as soon as a commercial break begins." They can even track tune-outs for a particular ad over time, every time it airs. Writes Dan:

So when viewers tune away from an ad, they are voting with their remote controls. These "votes" can be useful information, because they might tell us when viewers aren't finding certain ads interesting or relevant, and they bring us closer to understanding when TV ads are relevant to viewers.

Of course, they can't measure when viewers go to the kitchen to get a snack. But maybe there's a multiplier -- if a certain number of people change channel, x times that amount are ignoring the commercial altogether.

If Google can demonstrate to advertisers that fewer viewers tune out some ads over others, maybe advertisers can increase relevancy to particular demographics watching a particular program. Or at least demonstrate which ads are really boring.

So why hasn't anybody done this before? Maybe because broadcasters don't want advertisers to know how many people tune them out.

I'd still like to see how they determine which ads are relevant to newspaper readers, though...

Microsoft BrowseRank: Google killer?

Sorry, not likely.

I'm all for Microsoft and Yahoo figuring out how to become more competitive with Google search, and applaud their efforts -- as long as they do it separately. Combining the two search engines would have taken time and attention wasted on integrating existing products, rather than pursuing new ideas.

So Microsoft has announced a new algorithm in a paper titled "BrowseRank: Letting Users Vote for Page Importance."

Stephan Shankland at CNET says in his Digital Media column that "Microsoft thinks it can teach its rival a thing or two about searching the Internet."

But that's just the hyperbole journalists use in their ledes to capture attention. Kind of like describing something as a "Google Killer" in a headline.

The folks at Microsoft are fully aware that Google probably does this sort of thing already. Otherwise, why publish a paper describing how it works, allowing Google to adopt the new idea?

As Shankland notes in quoting a Google statement, "PageRank is just one of more than 200 signals we use to determine the ranking of a Web site."

But Microsoft's paper throws in a little hyperbole as well: "Experimental results show that BrowseRank can achieve better performance than existing methods, including PageRank."

The method uses browsing activity, rather than just links, to determine relevance. "The more visits of the page made by the users and the longer time periods spent by the users on the page, the more likely the page is important. We can leverage hundreds of millions of users' implicit voting on page importance"

And where does Microsoft get that data? From "an extremely large group of users under legal agreements with them," according to the paper.

Uh huh. "Extremely large." Well, there are a lot of people using Internet Explorer. But is Microsoft going to set up legal agreements with all of them, or just use the data?

Google, by virtue of its search engine, and probably things like Google Accelerator, routing traffic through its servers, also has an extremely large set of user browsing data. And you know that google is using every bit of user behavior it can to help determine relevancy.

The Microsoft paper is probably mainly intended to get publicity, to show the world it's spending its own considerable cash doing research, and hopes to get better than Google. And in that respect, the paper is a success.

Still, it's good to see that Microsoft is indeed working on the problem. It won't be a Google killer, because Google has spent many more years researching and refining its relevancy algorithms than anyone else, and catch-up is a tough game to play. It's kind of like all those years cyclists spent trying to catch Lance Armstrong powering up a mountain in the Tour de France.

But then, a big part of Google's advantage is its positive image. There have been studies showing that other search engines give just as good results. Everybody just assumes Google is the best.

A 2006 study by Intralink is one such example. I couldn't find that study. Must be gone. But a brief article in Cheap Hosting Directory said, "The most surprising discovery was how well MSN and Google performed in exact matches."

And in that case, a little positive PR for Microsoft is not a bad thing at all. Apparently, Ballmer's speech in Seattle didn't help.

If Steve Jobs is ill, does he have the right to keep the information private?

Rich Karlgaard at Fortune has a very thoughtful discussion over whether Steve Job's health is a private matter. If he is seriously ill, should he disclose the information because of the fact that people's pension funds and livelihoods are dependent upon him? Or is it callous to suggest that money trumps the right to privacy?

This is an extraordinarily difficult issue, about which I have many doubts. But in the end I have to say that in the modern age of health insurance, the right to privacy is a fallacy.The only benefit I can see for most people in keeping their illness a secret is in order to get medical coverage, in which case their insurance will be canceled if the insurer finds out you lied in your application.

Yes, certain illnesses, such as AIDS, carry a social stigma, and there is a benefit to the victims keeping the information to themselves. But the question with Mr. Jobs is whether his cancer has returned.

I lost my sister to cancer two years ago, and it was the most traumatic experience of my life. Thankfully, she had medical insurance. There was no secret about her illness, nor any thought of keeping the information from anyone who loved her, worked with her or in any way depended on her. What would she have gained? What would we have gained?

To those who say it's callous to insist that financial concerns take precedent, I have to ask, what does Mr. Jobs gain from keeping such a secret – assuming he's ill – besides the temporary financial health of his company?

I knew a powerful CEO who once kept the news of his cancer private simply because he was afraid valuable employees would bail out, worried that the company would fail. Of course, the company would have failed if he had died as well, but no one would have had the opportunity to protect their own interests -- or to show their loyalty to the CEO by standing by him.

There is a public good in such a public figure as Steve Jobs disclosing whether he is seriously ill. I fail to see the private good to him by keeping it a secret.

Microsoft: Those who can't do, advertise.

Ben Worthen, a blogger with the Wall Street Journal (you can read WSL blogs without being a subscriber) noted recently that Microsoft has decided to take on Apple. Not on design or ease of use, but on advertising.

“We’re drawing a line right here on this stage,” [Microsoft VP Brad Brooks] told a crowd of Microsoft business partners, according to a transcript. “You thought the sleeping giant was still sleeping? Well, we’ve woken up and it’s time to take our message forward,” he added. Brooks specifically mentioned Apple and that company’s “I’m a Mac, I’m a PC” ads.

Apparently, people still don't want to upgrade to Vista, and Microsoft wants to convince them that the OS ain't all bad.

Ed Bott at ZDNet noticed a flat-earth ad on Microsoft's home page.

Is this the best Microsoft can do? No, apparently this isn't the start of the new, hard-hitting ad campaign, reports Brooks.

Come on, Microsoft, this is a losing war. Microsoft battling it out with Apple on the ad front is kind of like ... well, Microsoft battling it out with Apple on product design.

It's a waste of money. I had a conversation with Scott Weiss, the CEO of IronPort Systems (now part of Cisco Systems) about how to be successful. One of his main points was the radical idea of giving your customers what they want. Product, design, service, packaging, customer support -- that's where you spend your money.

Said Weiss: "It's better to spend that marginal marketing dollar to provide service that creates free marketing through word of mouth. That kind of marketing you literally cannot buy."

Unfortunately, Microsoft's business strategy was to set standards and create monopolies and "barriers to entry." Somewhere in that strategy, the needs of the customer got lost.

Too many corporations forget that fact. If you want to do well, do well by your customers. Pay attention to their needs, even above your own. It's number one on Google's list of avoiding evil: Focus on the customer and all else will follow."

Corporate evil comes in many forms. Putting shareholders ahead of customers, creating products to maximize profits instead of filling customers' needs, sacrificing website usability for advertisers, sending manufacturing to China to save money and getting back toxic products -- these are all things a responsible corporation would do better to avoid.

In the end, you get back customer loyalty. And no advertising in the world can buy that.

Not even Apple's.

Carl Icahn posts

Since Carl Icahn has settled with Yahoo, he's decided to blog about the issues.

The highlights of the settlement, he says, aside from getting board seats for himself and two others of his choosing:

I will be offered the chance to serve on any committee established to consider material transactions out of the ordinary course of business.

and

"..the board has agreed in the settlement agreement that any meaningful transaction, including the strategy in dealing with that transaction, will be fully discussed with the entire board before any final decision is made."

In other words, he gets to lobby from the inside the next time a buyout offer is made. I hope it doesn't come to that. But you know that the only reason he's there is to get a buyout at a price that makes him a profit, whether or not the deal is actually good for Yahoo.

Otherwise, how could he possibly have supported the idea of crippling Yahoo by supporting Microsoft's bid to buy Yahoo's search business and leave the rest?

Face it: Icahn is a corporate raider. He is out for one thing -- flipping the stock after getting a buyout. He doesn't know how to run a tech company for the long haul.

Carl Icahn is to corporate management what Attila the Hun was to international relations.

But we'll be hearing more from Carl the Blogger.

Now that the proxy fight with Yahoo is over, certain constraints concerning my views about corporate governance have been removed and I intend to communicate more frequently in the near future.

I can't wait.

The Earthlink deathwatch

I don't see how Earthlink can possibly survive much longer doing business as it does.

Several months ago I wrote about my experience with Earthlink as a DSL provider. There have been several comments from people sharing their experiences -- pretty much exactly like mine. Tech support can't help. they put you on hold, then hang up. I once determinedly sat on hold for an hour before I got a message saying that I had been holding too long and they had to go serve other customers!

I wonder what customers these were.

Phil just posted a comment asking what ended up happening, so I thought I'd write a follow up.

The short answer: Absolutely nothing. Even after I filed a complaint with the FCC, all that happened is that I got bumped up to a higher level of management, which was equally unhelpful. After a while, I got a letter from the FCC saying that it understood my problem had been resolved. I wrote back and said it had not.

One of the Earthlink technicians decided that the problem was in my modem. They sent me a new one, then sent me a bill for it. the new modem made absolutley no difference to the problems I had, static on the phone line and on-and-off DSL.

They sent another technician out, who claimed, once again, that the problem had to be with my internal phone lines. I replaced all my lines with brand new ones from the main box to the walls, installed new jacks in the walls and bought a new phone just to make sure.

Guess what? Nothing changed.

I finally quit Earthlink and went with another provider. All of a sudden, my problems mysteriously went away!

I had to argue three more times with people in billing and threaten to go to the FCC again to get Earthlink to agree to cancel my last bill, the charge for the new modem, and the early termination fee. But I did convince them.

Anybody want a free, brand new Earthlink modem?

The really funny part is that a month after i quit, I got a phone call from Earthlink conducting a survey about customer service. I spent 20 minutes rating the company on a scale of 0 to 10. How was the service? 0. Would you recommend Earthlink to friends? 0. How likely are you to use Earthlink in the future? 0.

Does anybody actually pay attention to the survey results?

It is absolutely impossible for a company to stay in business this way, especially since my experience was far from unique. Before it goes out of business, file a complaint with the FCC and then cancel your service. You can do it easily online.

I got another letter from the FCC yesterday saying, once again, that they understand my issues were resolved. It's easy to just let it go, but I guess I'd better respond. Somebody needs to know how incompetent this company is.

Yahoo: Time for a rebound!

Good news from Yahoo has been as rare as teeth on a duck-billed platypus in recent months. The company releases earnings today. I'm hoping for signs of a turnaround.

Motley Fool is pessimistic:

Fellow Fool and CAPS All-Star Seth Jayson doesn't like the current management team here: "The boneheads running Yahoo! are looking to keep their jobs rather than do what's right for shareholders. They'll continue to screw up royal, and shareholders will eventually end up having to take a much cheaper buyout offer down the road."

Yeah, well all Jayson is looking for is short-term gains for shareholders that a buyout would bring. That says nothing about whether Yahoo would be a better company post-merger. I call that a bonehead opinion.

And who says management is only screwing up? Yang hasn't had much of a chance to make things better yet, but there are some good signs. First, Yahoo and Microsoft both gained share against Google in June, according to comScore. Yahoo gained 0.3 of a percentage point to captre a 20.9% share, while Microsoft's share rose 0.7 points to 9.2 percent.

But Microsoft's gain probably came because it's paying people to use its search engine. I call that a desperation move.

Yahoo's gain is organic. Plus, its ad deal with Google should start showing some results. It's holding off on implementing that pending antitrust review, but perhaps there was a small boost from their test of the program.

CNET notes that Yahoo is more exposed to the economic downturn than Google, because Yahoo relies more on display ads, which advertisers might cut back in a recession because you don't have demonstrable results. And CNET is right.

But CNET actually quotes an analyst who admits that long-term, the Google ad deal is probably better for Yahoo than letting Microsoft take its search business. Cowen analyst Jim Friedland:

"We believe that the Google partnership is the best choice for Yahoo, barring an acquisition of the entire company."

Neither is Google's disappointing quarter an indication of trouble in adsearchland. Lower interest income and legal expenses defending YouTube hurt the bottom line.

Analysts expect a 13% drop in earnings at Yahoo, so all it has to do is beat that estimate for things to look good. Wouldn't a stock price jump be nice?

With a compromise with Icahn in the works, maybe Yahoo can start getting some respect. a decent quarter among all the doom and gloom could do a lot to help, and perhaps put Microsoft -- which has been working hard to hit Yahoo's stock price in hopes of a fire sale -- in its place.

Go, Jerry!

UPDATE:

Yahoo just reported net revenues of $1.346 billion. FactSet Research says Wall Street was looking for $1.38 billion.Net income was $131 million or $0.09 per share. Street was looking for $0.10 per share. A little short there, Jerry.

Late in the day the stock had rallied a little after an early decline, although it didn't make it back to yesterday's close of $21.67.  In after hours trading as of 4:52 pm EDT, it had gained $0.20 to $21.60.

Is Google becoming the new Microsoft?

Just how powerful can Google become, and what are the consequences?

These questions deserve attention, and are getting a lot lately -- some hyperbolic, some not.

Bill Snyder at Infoworld is raising the alarm because of the potential advertising tie-up between Yahoo and Google. In "Warning: Google is becoming Microsoft's evil twin" he writes,

" Sergey Brin and the gang that will do no harm have learned the worst possible lesson from Microsoft: build a monopoly and they will come -- because they have to.

Being naive, and not owning any Microsoft stock as Snyder does, I thought that Google was offering its services to Yahoo in order to avoid the necessity that Microsoft buys Yahoo.

The deal with Google, he says,

" is bad for business, bad for consumers, and bad for IT. It will raise Web advertising rates by more than 20 percent. It ought to be stopped.

Where did that number come from? SearchIgnite, which helps advertisers make their ads more productive. Searchignite compared the cost per click of keywords, and found that they cost 22% more at Google than at Yahoo. It's a kangaroo's leap to say that the deal will automatically raise Yahoo's prices by that much.

This analysis ignores the fact that advertisers bid on the price per click, which sets them at a fair market value. If you're not getting results to justify the price, you bid lower. The Google ads have higher payoff, so the bids are higher.

There are two scenarios: Either bid to just place your ads on Yahoo and get them cheaper, or bid on Google's entire collection of AdSense sites and pay more. This isn't exactly Microsoftean evil.

Neither is this on a par with the kind of tactics Microsoft used to get itself into trouble. Microsoft didn't let PC makers bid on their software, it set prices. If the customers dared to install competitors' software on their PCs, Microsoft charged them higher prices for Windows.

The comparison of this deal to Microsoft's tactics is just silly.

And, I may note, it was Google News that led me to Snyder's article. That's not the tactic of an evil monopolist.

Nevertheless, prices may go up with this deal. When there's only one auction to bid under, prices are likely to increase. That's a concern. Competition is good and as necessary as a well-armed militia.

Here's the solution I propose. Mimic Google, don't fight it. Microsoft and Yahoo should make sure their ad systems link to the most relevant content, are not obnoxious and rise to the top of the list if they're effective, regardless of the price people pay for them.

Then Google won't be so much of a monopoly.

(And I don't own stock in any company. Period.)

In defense of Jerry Yang

Andrew Ross Sorkin at the New York Times spent some of his time at the Sun Valley conference talking to people about Microsoft, Yahoo and Carl Icahn. He wrote a nice piece in defense of Jerry Yang. It's worth reading.

Apparently the "moguls" at the conference were appalled at the "deal" that Microsoft's Steve Ballmer and Icahn worked out behind the scenes at the conference.

The new offer calls for Microsoft to buy Yahoo's search business and for Icahn to take over the rest of the company. According to Sorkin:

“At this point, it’s embarrassing,” one technology C.E.O. whispered just out of earshot of Bill Gates. “Ballmer should either buy the company or forget it.”

Yang says he has been willing to sell the company all along at the right price, but feels Microsoft hasn't been negotiating in good faith. It's just trying to kill the stock in order to pick it up at a fire sale.

Sorkin is fond of quoting anonymous moguls, and this one agrees with Yang's assessment:

“Steve is playing Carl to the hilt,” said one rival mogul, drinking a Scotch and smoking a cigar after midnight. “If Carl gets the board, Steve will buy Yahoo for a song — or watch Yahoo die on Carl’s watch.”

That sounds like something Larry Ellison or Scott McNealy would say. Are they at the conference?

Microsoft to buy AOL

It was March 2007 when I first predicted Microsoft would buy Yahoo.

I was wrong. Again. Outright acquisition seems (rightly) out of the picture now.

Kara Swisher at AllThingsD has a better idea. Perhaps Microsoft and Yahoo will turn their attention to buying AOL. I like that idea. Microsoft will win any bidding war.

What does Microsoft get from buying Yahoo's search business, anyway, besides the indelible impression that it can't create a viable search engine to save its wealthy ass? It gets traffic. It gets eyeballs to blind with its own flashing ads. Then it could show Google, boy.

What does it get if it buys AOL? Eyeballs that are currently seeing Google search results and Google ads.

Microsoft could then prove its software machismo by replacing Google's search engine at AOL with MSN's (will anyone notice the difference?) and switching the site over to Microsoft ads. It gets twice the punch for that purchase.

Of course, AOL is damaged goods, because the celebrated wedding with Time Warner didn't work out so well. But it would be a better fit with Microsoft and MSN. MSN could start losing money twice as fast!

Did I say that? No, I mean that Microsoft could finally make MSN profitable with all those new eyeballs staring at its award-winning content and ads.

Come on, Ballmer. Go for it.