The bigger Google grows, the more it seems that like every large company it's products now have to pay a strategy tax which may end up shortchanging their users. If you are unfamiliar with the term, you should read Dave Winer's excellent essay Strategy Tax which is excerpted below

Ben explained that sometimes products developed inside a company such as Microsoft have to accept constraints that go against competitiveness, or might displease users, in order to further the cause of another product. I recognized the concept but had never heard the term.

An example. Consider a company that develops both a Web browser and a word processor. The product team for the browser might get lots of input from users saying "We'd like a better editor in the browser." It would be natural to give the users what they want. So they put the feature in their project plan, but when it comes up for review, the CEO shoots them down. "They should use the word processor," he says.

Another example. An electronics company makes a portable music player, and also owns a recording studio. The team working on the music player knows that there's a big market for players that work with a format that's popular with computer users, but lacks sophisticated copy protection. They project huge sales. But when the product comes up for review, the CEO shoots them down. "We'll lose sales of our music," he says.

Before I read this essay I thought this was a problem unique to Microsoft and also thought that I was being astute in observing the trend at the company when in truth the term was part of the cultural knowledge of the company while I was still in programming diapers. Over time, it has become clear to me that this is a problem that affects any business endeavor where different product units either rely on each other or indirectly compete with each other [assuming that the company isn't dysfunctional enough to have products that directly compete against each other]. Below are three examples of how the strategy tax is affecting Google, all of which are observations by other bloggers which I've noticed myself but refrained from mentioning since I work for a competitor and it would have come off as sour grapes.

Disincentive to Improve Search Due to Ad Business

In his post Good Luck Jimmy, Dave Winer writes

Google is repeating the pattern of the previous generation of search engines (Alta Vista, Infoseek) were doing when Google zigged to their zag, so successfully. Today, Google is fattening up and spreading out, going after Microsoft in productivity apps, chasing the TV networks with YouTube. Etc etc. Today search is only one of the things Google is doing, and it may not be the most important thing.

Today Google's profits come from ads, and that business gives them a reason to keep search weak. They want you to do a lot of searching to find what you're looking for -- and the stuff they find for you for free is competing with the stuff they make money on. So Google actually has a disincentive to make search better.

A few months ago, I used to get into regular debates with Matt Augustine who argued that the fact that companies like Google make so much money from search advertising seems like a bug in the system. Matt would argue that if search engines were really that good at finding what we want, we would never have to click on the results they had been paid to show us unless we were deceived into doing so.

This seems to put a 'glass ceiling' on how good the search engine can be because you don't want people to stop clicking on ads when you make billions of dollars a year from them doing so.

Promoting Google Services at the Expense of the Integrity of Search Results and it's Advertisers

Blake Ross has a blog post entitled Tip: Trust is hard to gain, easy to lose where he writes

But Google lost me today, and it didn’t take much:

Google is now displaying “tips” that point searchers to Google Calendar, Blogger and Picasa for any search phrase that includes “calendar” (e.g. Yahoo calendar), “blog” and “photo sharing,” respectively. This is clearly bad for competitors, and it’s also a bad sign for Google. But I generally support anything that benefits users, including monopolistic packaging. I believe, for instance, that shipping Internet Explorer with Windows was a good move. So why are tips bad for users?
The tips are different—and bad for users—because the services they recommend are not the best in their class. If Google wants to make it faster and easier for users to manage events, create a blog or share photos, it could do what it does when you search GOOG: link to the best services. To prevent Google from being the gatekeeper, the company could identify the services algorithmically. But if that sounds familiar, perhaps that’s because Google already works that way. After all, Google is predicated on the idea that the democratic structure of the Web will push the cream to the top. Search for “photo sharing” and you should already get the highest quality services. According to Google, Picasa is not one of them.
While advertisers compete to be first in a string of lookalike ads that are often shunted to the side, Google now determines the precise position and appearance of ads tips that are not subject to any of the same rules. Its ads get icons while others don’t, and if you think that’s small potatoes, you are not an advertiser: images boost clickthrough. Google can make a Picasa ad say “Easier to use than Kodak,” but Kodak cannot create an ad that reads “Easier to use than Picasa.” And the kicker: neither the highest quality ads nor the highest quality search results can replace these tips.

The "strategy tax" here is being paid by the search engine and advertising groups at Google. To carry along Google services that Blake points out are not best in class, Google is foregoing ad dollars from a number of lucrative keywords and causing distrust in the search engine by the very power users upon whose backs it rose to fame in the first place. Google used to brag about how unlike other search engines, they don't use misleading ads that people can confuse for search results. However I tend to agree with the last statement in Blake's post

Perhaps the most nefarious aspect of this feature is how it operates within our collective blind spots. Advertisers are happy that Google no longer invades the canonical Ad Results. Technology purists continue to see untainted Search Results. But does my mother make that distinction? How much does a result have to look like a Result to cross the line?

Artificially Promoting it's Products in Search Results

From a comment highlighted in the post Google's Silent Monopoly Redux (Google Responds - Issues Public Statement) which states

But type in "maps". Google is again first. Ahead of Mapquest. Ahead of Yahoo maps. Yahoo also has backlinks out the ying yang. So why is it third? And mapquest has been around forever.. I'm sure there are more links to than to the URL, simply because the millions of web pages that linked their directions to Mapquest from 1996 to 2004 didn't all rush out and change all their links to Google maps in February of 2005 (when it was released), even if Google's is a better product.

Next, try "mail". Despite the fact that Yahoo mail has been around forever, and has all sorts of links, and that Hotmail has also been around forever, Gmail still manages to come up first.

And the most interesting thing about this particular keyword? The word "mail" by itself doesn't even appear on the page! The words gmail, webmail, and email appear. But not "mail". At least on the Yahoo page, the word "mail" does indeed appear. Yet Google still manages to rank ahead of Yahoo.

Finally, try "answers". Yes, comes up second, rather than first. But comes in third! Is the Google Answers site really getting that many more links than Yahoo's? Especially in light of the fact that Google recently decided to kill it, because almost no one was using it, while Yahoo's usage (and therefore also linkage, no doubt) are skyrocketing?

This claim was actually the most interesting to me since Google is very adamant about the integrity of their search results and claims we don’t accept payment for inclusion in our index, nor do we manipulate search results by hand. I tried a number of these queries myself and was pretty shocked by the results especially when it came to "mail". Here are some screenshots that illustrate the point

1. Search results for "mail" on Google

2. Number of links to (also the same as according to Google

3. Number of links to according to Google

It's hard to imagine any objective metric that should make Gmail show up ahead of Yahoo! Mail in a search for the word "mail". Of course, this doesn't mean that Google is tampering with search results "by hand". Their algorithm can simply have allowances to rank sites in their domain or linked from their domain higher without having to actually sully their hands by tweaking individual results by hand. Still, if Google is how the world finds information and we are increasingly being pointed to information that financially benefits Google, doesn't that taint the much vaunted claim of the integrity of their search results even if it is being done in an automated manner?


Categories: Competitors/Web Companies
Tracked by: [Pingback] [Pingback] [Pingback] [Pingback] [Pingback] [Pingback] [Pingback] [Pingback] [Pingback] [Pingback] [Pingback] [Pingback] [Pingback] [Pingback] [Pingback] [Pingback] [Pingback] [Pingback] [Pingback] [Pingback] [Pingback] [Pingback] [Pingback] [Pingback] [Pingback] [Pingback]
"Ab belt" (Ab belt) [Trackback] [Pingback] [Pingback] [Pingback] [Pingback] [Pingback] [Pingback]

Wednesday, December 27, 2006 7:33:47 PM (GMT Standard Time, UTC+00:00)
This is a great read. I can see many correlations between this and Amazon. But they aren't nearly as visible to an outside observer because we aren't battling with Google or Microsoft.

One thing though - here is the URL I use to access gmail - It has mail in there twice. I agree that Yahoo! Mail has to be more popular just because of age, and it is weird that the number 1 link is "gmail" since they accept "mail" as well...
Wednesday, December 27, 2006 9:04:16 PM (GMT Standard Time, UTC+00:00)
Is it conceivable to you that Google marketing manager's play fair when bidding on keywords and that Gmail showing up first in search results could be legitimate (do we even need to point out to you that Google values link quality over link qunatity?)?
Wednesday, December 27, 2006 9:29:15 PM (GMT Standard Time, UTC+00:00)
Even if Google's marketing folks were playing fair [for some definition of "playing fair"] when they were bidding on Adwords, any pretense of that is gone now that they have replaced the ads with "Tips" that only pitch Google services and are more prominent than the ads.

As for Gmail showing up higher in search results than sites that have been around for several years longer, have an order of magnitude more users and an order of magnitude more links to them...excuse me if my ability to suspend my disbelief fails to come up with a plausible reason that GMail is somehow more 'relevant' than Yahoo Mail or Hotmail.

Like you, I even entertained the thought that A-list bloggers with high pagerank linking more to GMail than Yahoo! Mail or Hotmail could be a factor so I performed some Technorati searches sorted by authority for and, the results showed that Y! Mail seems to be holding its own against GMail in references by A-list bloggers (although Hotmail isn't). Any other ideas?
Thursday, December 28, 2006 1:15:44 AM (GMT Standard Time, UTC+00:00)
Funny thing though Dare, doing the same on Gmail is only a couple down - sure it's behind Yahoo! but Hotmail doesn't make a showing and not even on the second page where we hit Live Mail.

I guess it shows that you guys are doing the right thing as far as not skewing results, but it also shows that Gmail remains fairly high up there.

So I guess that Gmail is more relevant somehow than Hotmail, even if like you say it's A-list blogger or something.

Maybe you could do some comparisons with other search engines?


(apologies if this posts twice)
Andrew Tobin
Thursday, December 28, 2006 2:37:55 AM (GMT Standard Time, UTC+00:00)
On the subject of strategy taxes and internal conflicts, in about 1981 I used to work for Texas Instruments just as they had decided to rid themselves of strategy taxes. They had the problem that they sold chips but also sold products that used the chips: they had a large laser printer business, a PC business, micro-mainframe business and their LISP machines (including their Explorer computers: multiple CPU boards running direct LISP or XENIX in any mix.)

So they divested themselves or closed down all the product businesses, and concentrated on their core competency, which was taking advantage of their IP and chipmaking. It was a really horrible experience for the workers: amazing products like the Explorer were killed off, computing plunged back into the dark ages...err I overstate. (It had so many features that MS still cannot get right: near instant-on for example.)

So MS is certainly not the only one. Another example of the futility of internal competition is with research labs with large corporations. I have been told about IBM research in the last decade, but I imagine the same dynamic holds in other large companies. The problem is that that only a fraction of research gets productized because it may conflict with current strategies; if it does conflict, there is little chance of taking the techology outside the company.
Rick Jelliffe
Friday, December 29, 2006 5:23:29 AM (GMT Standard Time, UTC+00:00)
Just another point of comparison - if you search for "email". Yahoo Mail comes out ahead.

Friday, December 29, 2006 8:24:15 PM (GMT Standard Time, UTC+00:00)
Here in Germany Google Search returns
for "photo sharing": 1. Flickr, 2. DropShots, 3.
(for language=English) and no Picasa on the 1st page
for "mail": 1. Yahoo (Germany), 2. Gmail, 3. 4. GMX, 5.
(for language=English)
1. Yahoo (Ge), 2. Wikipedia, 3. Lycos,
but then, as if I mispelled "mail":
"Ergebnisse anzeigen für: gmail" with
1., 2. Gmail (Wikipedia), 3.

I'm not yet sure what to make of this...

Tuesday, January 2, 2007 9:51:18 AM (GMT Standard Time, UTC+00:00)
I remember reading a paper that covered speeding up the computation of Page Rank by computing the intra-domain Page Rank and then combining it with the relative rank of the domain. With that approach, links to would add rank to and
Comments are closed.