Yahoo’s Poor Ad Targeting & Thoughts On Google-Yahoo

Previously, I wrote of how Yahoo’s recent press day proved disappointing to me. Time for part two of that: how the Google-Yahoo ad deal is supposed to help Yahoo. Oh, it might, but not for the reasons Yahoo put out — the idea it would somehow get advertisers it was missing. More about that below, as well as thoughts on Google’s auction that’s not an auction and whether the Google-Yahoo deal should be blocked.

As part of the press day presentation, Yahoo executive vice president Hilary Schneider showed how few ads Yahoo returned for a search on [red roses in birmingham alabama]. In contrast, Google’s search results page was loaded with ads. By partnering with Google, Yahoo would thus be able to supplement its own ads with these additional ones that it lacked.

Sure, Yahoo had fewer ads. But that’s not because it lacked advertisers that Google has. It’s because Yahoo’s ad targeting system is pretty lame.

Advertisers Use Yahoo; They Just Don’t Show Up

I went through the ten ads that Google was showing at the time for red roses in birmingham alabama, to understand what terms those advertisers might have been targeting. One way to tell this is to see what terms are bolded in the ads, when they appear. The list looked like this (the domain/advertiser is shown first, then the terms they were matching):

1. 1800flowers.com (nothing was bolded, so they were probably matching “roses” via a synonym like flowers)
2. delaflowers.com (red roses)
3. teleflora.com (birmingham)
4. proflowers.com (birmingham)
5. ftdflorists.com (birmingham)
6. justflowers.com (birmingham al)
7. doubleknockoutrose.com (rose)
8. birminghamflorist.net (birmingham)
9. localflowershop.com (birmingham)
10. frequentflowersinc.com (birmingham)

Next, I wanted to find out how many of these advertisers were actually advertising on Yahoo, since part of the argument has been that Yahoo might not have some of them. Turns out, all but one of them were. I discovered this by doing various searches related to flowers and roses, where I could see the same advertisers appearing.

Hmm. This told me Yahoo’s issue wasn’t that it lacked advertisers. It just isn’t getting their ads more broadly placed.

Google’s Broad Match Versus Yahoo’s Advanced Match

Over at Google, ads are placed by default against a wide range of terms automatically through what’s called “broad match.” So if I create an ad at Google, tell Google that I want to match the word “roses,” it will automatically have me match a variety of other terms related to roses, including “red roses in birmingham alabama.”

Just to be certain of this, I tested it out when writing up this article. In only a few minutes, after only listing the single word “roses” to target, thanks to broad match, I was ranking for “red roses in birmingham alabama.”

Yahoo has broad matching too. They call it “advanced match,” and I was really puzzled about why so many advertisers were failing to show in spite of advanced match almost certainly being used by many of them.

Remember that list above? These advertisers:

• teleflora.com
• ftdflorists.com
• birminghamflorist.net
• frequentflowersinc.com

Had the EXACT same ads on Yahoo as were running on Google, such as this one FTDflorists.com was using:

Flowers Birmingham
Official FTD Directory of Local
Florists. Shop online, Call, Visit.
www.FTDFlorists.com

Same ad, both places — same advertiser clearly targeting people looking for flowers in Birmingham, and that would include red roses. But Yahoo fails to get their ad showing, when they almost certainly were targeting either the words “flowers” or “roses?”

That took me back to testing, this time over at Yahoo. I made an ad and told that by default, advanced match would be used. I then put in the single word “roses,” sat back and waited.

My ad showed for “rose” and “roses.” It didn’t show for “red roses” much less “red roses in birmingham alabama.” Perhaps it might have shown for some other flower-related terms. But that’s beside the point — if Yahoo can’t target an ad for “roses” to “red roses in birmingham alabama,” it’s got serious issues.

OK, some kudos to Yahoo. If you target ads too broadly by default, you can have advertisers upset. Someone might not ship nationally, for example — so they want to target to a specific geographic area. But Google is not overwhelmed with complaints about things being targeted too broadly. This has been an issue in the past, and I’m not saying the issue is gone — only not a major one that I’ve heard about recently. If Google’s doing OK, why can’t Yahoo really go more broadly?

Another problem is that Yahoo does try to help advertisers go broadly in the account setup option by offering to suggest terms. So when I let it do that, I got this:

florist
red rose
flower arrangement
flower
flower delivery
send flower
flower online
fresh flower
white rose
flower shop
flower gift
bouquet
how to plant roses
order flower
corsage
holiday wreath
florist in
wedding flower
buy flower
funeral flower
flower bouquet
flower roses
local florist
gift basket
dried flower
flower centerpiece
buy flower online
thank you flower
birthday flower
rose bouquet
online florist
gift
rose send
florist shop
order flower online
send flower online
cheap flower
sympathy flower
online flower delivery
flower delivered
floral
florist online
plant
florist flower
mother day flower
valentine flower
anniversary flower
gift idea
dozen roses
buy roses

Sure, that would get my ad out in more places, but then again given how closely Yahoo also seems to stick with selected words, I still might now show up for a variety of other important searches. In contrast, I were to target the word “flowers” at Google, it’s going to get me out for any search using that word along with synonyms and related terms.

Important side note. I noticed that when I repeated certain searches on Google, some or all the ads eventually went away. To see them again, I’d have to shift to a different browser (changing from Firefox to Internet Explorer, for example). And then sometimes after only a few minutes, the ads would go away yet again, forcing me to hit a third browser to bring them back. I’ve never seen this happen before, and it might very well be common knowledge to those deep in the paid ad game that Google acts this way. But it was new to me.

NOTE: Reader Matthew Gatley says this is probably due to the new frequency capping feature Google recently announced. However, that seems limited to the content network, rather than search-targeted ads. I’ll be checking on this.

Back to Yahoo, I pointed out in questions after the press day presentations were done that the example it gave wasn’t proving that Yahoo lacked advertisers but only that it had sucky targeting options for those advertisers. Yahoo acknowledged that the broad matching would have to be improved and that the deal with Google allows them time to work on this.

The World’s Weirdest “Auction”

That brings me back to the debate I see which is getting renewed once again, whether or not an ad deal with Google and Yahoo is something to be feared. I’ll dive into this more with a future article, but the key concern to those opposed is the idea that Google could set prices because it would have an estimated 90 percent share of the paid ad space in the United States.

The “auction” argument never goes well with me, given how heavily Google manipulates pricing anyway. Google manipulates pricing? Of course. The mystery “quality score” factor (and see here and here) means some advertisers will pay much more than others to appear (and conversely, it should be noted, some will pay far less).

To explain what a mockery this makes of the auction process, imagine Google’s selling a painting. You’ve got two people in the front row bidding for it. One’s an older woman, and to make it all stereotypical, she’s there with her pearls around her neck and little poodle in her lap. Next to her, we’ve got some unshaven guy in ripped jeans swigging a bottle of Thunderbird. The bids come in:

• Unshaven guy: $8 million
• Pearl & Poodle Lady: $6 million

Now Google looks at the bids and says “sorry” to the unshaven guy, but because Google doesn’t like the way he looks, $8 million won’t cut it. If he wants the painting, he needs to pay more to prove his “quality,” and they give him a price of $35 million. He walks away, and Pearl Lady wins.

That’s an auction?

Who “Sets” Millions Of Prices?

Google’s algorithms are already setting prices, and the advertising groups that I’ve seen speaking out about this “new” concern that might happen as part of a Yahoo tie-up don’t seem to understand that so many prices are set on a constantly changing basis that Yahoo’s not really going to make a difference. How are they going to collude — on millions of terms, where the prices differ for each term for each and every advertiser?

Sigh. It’s like some of them think Google’s got a single rate card that it will compare to Yahoo’s rate card and then the both agree to mark things up. It doesn’t work that way, folks.

Moreover, let’s assume the Google-Yahoo deal doesn’t go through. There’s an excellent chance then that Microsoft will eventually buy Yahoo, say in 6-12 months. That means further uncertainty for both companies, followed by a 6-12 month approval process. And then add in another 6-12 months for Microsoft to digest Yahoo.

Google’s just going to storm along on the search front during this time, getting stronger and stronger. Even worse, Microsoft — sad to say — has only proven so far that it makes a mess of search. The entire Microsoft Windows Live branding problem doesn’t just go away. It is not guaranteed that Microsoft somehow will gain share by getting Yahoo. A Microsoft purchase might actually weaken the company’s prospects. As a result, Google might hit 90 percent share regardless of whether a Yahoo deal happens.

Can Google Be Blocked Regardless Of Yahoo?

When you understand some of these fundamentals, this seems to be the obvious questions to ask or examine:

What percentage of the paid search market (if any) does Google have to have before it is forced to be regulated?

70 percent? 80 percent? 90 percent? Or is it fair game for it to have as much share as it can get, as long as it has not shown to have acted in an anti-competitive manner?

A Google-Yahoo deal seems a side-show in answering this. If folks are certain Google will have too much power over pricing, there’s an excellent argument it has this already. So let’s not get the argument distracted over Yahoo.

Make no mistake — I have unease with Google’s continued growth. Somewhat similarly to Michael Arrington today at TechCrunch, I have issues where Google intersects with publishing (see my Google: Master Of Closing The Loop? and Deconstructing Google: Chapter 4, After The Google Antitrust Breakup articles for more on this).

I’m working on a new article covering some of these issues, and part of it asks whether Google should have simply walked away from partnering with Yahoo, to avoid the type of fallout happening. I had a chance to put this to Google cofounder Sergey Brin last week. While he understood the concerns, I also thought he had an excellent response:

“If I have a restaurant, am I supposed to not let some people eat in it?”

Remember, for years it was the norm for companies to outsource search. When companies like Lycos dropped out to be “powered” by others, no one started worrying about whether some other company should be “allowed” to provide those services.

Sorry, You Have To Use Microsoft?

In light of this, it seems somewhat absurd that Yahoo shouldn’t be allowed to do what it wants. I mean, it could decide to close its entire search operation tomorrow — both paid and unpaid — and say that it was out of that game. If it did this, would it actually be prevented from using Google in favor of Microsoft, even if Microsoft was likely to provide it with lower quality search results and less monetization?

To me, that’s sort of like wanting to buy a car, going to a Toyota dealership to get a Prius but being told you can’t have one, since the government has decided those have sold too well. In the name of competition, you have to instead buy some GM hybrid for twice the Prius price.

To bring it back to Yahoo, I’ve previously written that there’s an excellent chance that the deal with Google might allow Yahoo the time it needs to improve itself as a competitor to Google. I also have no doubt that Google wants that to happen — that in the Google mindset that’s strange to some, they do want competition (though they’d rather it be Yahoo). Yahoo’s certainly in a much stronger position than Microsoft to fight Google on the search front.

But still — after seeing such a fundamental failure of ad targeting, I have worries about whether Yahoo really will use the time to set things right. And part of me wishes Sergey had indeed told Yahoo that it couldn’t dine at his table. And part of me still doesn’t see why this might ultimately end of with Microsoft essentially a gift of Yahoo. Wish I had all the answers.