Barry Diller on search engine optimization

I dislike quoting a secondary source but I’m unable to connect to the All Things Digital Website as I write this so I’ll have to use a secondary source. Keep in mind that my information is only as good as my source, so I respectfully reserve the right to change my mind on anything that follows.

The Barron’s Blog reports that Barry Diller shared his view on IAC’s future at All Things Digital 2008 (ATD is an annual conference of “technology industry elite” — CEOs and such folk — so it generates a fair amount of press on a variety of topics).

Diller seems to be defending Ask’s poor search market share by emphasizing that Ask wants to deal with question queries. Nonetheless he is waiting for the day when Google slips up and then maybe Ask can capitalize on that mistake. I’ve never actually seen a top search engine tumble to a same-generation rival (then again, I’ve only seen search engines shuffle their rankings a few times). Based on past experience (which provides no guarantees for the future) I would expect Google to lose its dominant position to a next-generation search resource.

Diller also explains that he is splitting IAC into five companies so that each of the five logical groups can emphasize their own strengths. The new IAC will be the Internet venture, focusing on search and other Internet initiatives. The problem is that the collective IAC properties have always possessed unrealized potential. Ask could have significantly increased its market share if it had spread its search box across all the IAC properties (or at least built a link into those properties’ Web sites’ front pages). Ask would also have to ensure that each of those sites was fully indexed.

The lack of adequate site indexing is where Ask falls down on the job. It’s got a great search technology that provides high quality results but search engines have to look beyond what is most popular. If you used just citation-based analysis to do scientific research you’d undoubtedly fall behind current scientific thought quickly because older research papers have far more citations than newer papers. It may take 10-20 years for scientific ideas to be validated and win general acceptance. The transitional between one idea and another within the scientific fields is usually fairly gradual and requires a lot of discussion.

The World Wide Web, however, is not a bed of scientific literature. It’s the repository for everything that interests us: our thoughts, our dreams, our fantasies, our misguided or mistaken beliefs, our school term papers — the Web holds everything that humankind can possibly store electronically. That huge archive needs search in the worst way possible because it’s way too big for any one of us to sift through it all. And today’s search generally sucks at indexing all the content on the Web, much less helping us find what we are looking for. So it doesn’t help Ask’s situation that they refuse to index the majority of the Web.

This is the problem with over-reliance upon a specific methodology. Ask, like Google, wants to validate the process that it has championed but doing so creates a conflict of interest. It is not in the searchers’ best interest for sites with the most expert-hub relationships or the most PageRank to be shown first (or exclusively). Searchers quite often want to find content that the search engines intentionally bury or ignore. I know I do and I know that people who use site search are looking for things they cannot find in Web search.

Diller says he wants to increase queries at Ask (thus making market share irrelevant to Ask’s potential). To a certain extent I agree with his philosophy, but in other ways I disagree with him. Today’s search market share metrics are completely inadequate because they focus too much on number of queries and not enough (or at all) on other factors. Google’s real market share is not nearly as high as the metric companies report because they don’t look at the whole picture. Ask’s market share, however, probably is only slightly larger than the metric companies report because Diller has done just about everything he can to ensure Ask’s failure as a search engine.

I was both amused and disappointed to find that Mr. Diller feels his company is expert at search optimization. They buy other companies and improve their search performance — apparently in Google, since Ask doesn’t drive that much traffic to Web sites. There is far more to search engine optimization than just getting good referrals from Google, Yahoo!, and Microsoft. The search engineers need to engage in their own optimization, and Ask desperately needs to be optimized in ways that it fears to consider.

Ask is fast becoming a much beloved relic of the past. Old school SEOs like me will praise Ask for its technological accomplishments but we will also shake our heads over its timid marketing objectives and initiatives. In more knowledgeable and capable hands Ask could have become so much more. So when Diller says that Microsoft’s proposal to buy Yahoo! was a good move, he speaks with no credibility.

Microsoft has less value as a search engine because it possesses less market share than either Yahoo! or Google. Buying Yahoo! won’t give Microsoft more market share in the long term because Microsoft’s marketing initiatives have failed. Failed leadership doesn’t turn itself around by finding new followers. All that Yahoo! has accomplished and held onto would be at risk of vanishing in yet another search engine brand collapse (and history teaches us that search brands don’t survive buyouts). In a year or two Microsoft would continue to trail Google, there would be no Yahoo!, and everyone on Wall Street would be saying, “Hey, what happened to those advertising revenues we were all focused on?”

You cannot combine search engines because of their advertising revenues. No matter how strong the numbers look when you add them up on paper, you have to have a strategy for strengthening their combined search market share and neither Microsoft nor Yahoo! has put forth such a strategy. Maybe for competitive reasons they felt it best to say nothing on the issue but searchers have no compelling reason to stay with either Microsoft or Yahoo! if one brand prevails over the other.

Steve Ballmer did promise that both brands would be kept alive while publicly drooling over the prospect of bringing Yahoo!’s braintrust in-house. While expressing no faith or confidence in his own technological resources, Ballmer could do no better than to follow Barry Diller’s lead in suggesting that the solution to Microsoft’s problem was to buy someone who was doing better than Microsoft. While that strategy has typically worked in the past for Microsoft with spreadsheets and other software, Web search is an entirely different world. You cannot bundle the Web you want to prevail with every PC that is sold (a strategy Microsoft attempted to execute that was successfully blocked).

In Web search you cannot leverage brand names. Yahoo! tried that with Altavista and other search properties it bought and those brands are now pretty much worthless.

In Web search you cannot leverage hardware and retail contracts. Both Microsoft and Google have tried that and neither has enjoyed any success.

In Web search you cannot buy your way to success. Barry Diller and the old IAC tried that and they have meandered along touting the efficiencies of their search engine that only a few tens of millions of people use each month.

In Web search you only have one thing to leverage: Web sites. Microsoft and Yahoo! do that better than either Google or Ask, and that is pretty much why Microsoft and Yahoo! have achieved as much as they have in the wake of all the mistakes they have made. Google didn’t become the dominant search engine because people at Google did things right. They became the dominant search engine because other search engines screwed up and the searching public didn’t wait around for those services to fix their problems.

Although search metrics occasionally pay homage to the concept of “core audiences” that are intensely loyal to each major search engine the truth is that searchers are fickle. They won’t stay with Google one moment longer than it takes them to realize that there is a search tool out there that serves their interests better. Google seems to get that. Yahoo! and Microsoft seem to be struggling with the concept. I don’t think Barry Diller has any hope of approaching reality on that level.

Search engine optimization is founded upon the concept that you have a Web site (not a brand) that is worth using, that should be visible to everyone, and that will accrue value over time as more people find it. It doesn’t matter if you’re trying to get into the index or if you’re trying to BE the index. Ask has failed to pass the test on both fronts.

IAC may really have some good SEOs on staff. Unfortunately, it doesn’t have a good marketer at the helm and IAC needs good, effective marketing leadership more than anything else. Being good at buying companies doesn’t make you good at building market share. If you don’t know how to leverage the resources you already have at your disposal you won’t do any better by acquiring new resources.

Ask, Microsoft, and Yahoo! all have potential they don’t know how to realize because they don’t believe in themselves. Yahoo! was ridiculed for publicly saying they were happy to be the number 2 search engine. Microsoft has revealed it has no faith in its search technology staff. Ask has repeatedly wallowed in its choice to NOT index the entire Web and broaden its appeal to searchers.

Any one of these three search engines could significantly shake up the search industry by adopting a new technology, a new algorithmic principle, that complements its core indexing strategy. Google will continue to roll out PageRank-link fluff services and monetize them not because PageRank works well (it’s horribly incapable of producing good quality search results) but because no one else has the wit to challenge them in a way that appeals to searchers.

I don’t expect Jimmy Wales’ social search concept to topple Google, either. After all, Google’s artificial PageRank mechanisms is social search (people voting for Web sites with links). Google can refine that process with thousands of filters and special tweaks.

Search needs to look at the Web differently. What we have today is getting old. The searcher needs to get closer to ALL the data and not just the data that some search engine deems is “safe”, “appropriate”, or “good enough”.

In that respect Barry Diller and IAC have good company. They’re no worse off than the people at Microsoft, Google, and Yahoo!. The one inhibiting factor to innovation is success because people almost always sit on their pet theories and try to make them work against all odds.

Albert Einstein lost his relevance to physics about 25 years before he died.

Steven Hawking’s revolutionary ideas were challenged and discarded by the astrophysics community decades before he, too, obliquely conceded that he probably got it wrong.

The greatest scientific minds of the 20th century could not prevail against the slow methodical progression of science and technology. There are too few scientific minds at Ask, Google, Microsoft, and Yahoo! to hold back the tides of time. Google has done a pretty good job of scooping up Ph.D.’s by all accounts, but by indoctrinating every employee in the beauty of PageRank and its Holy Hand Grenade of Bullshit Google has imploded its resources, opening up an opportunity for other minds to dream big and look beyond yesterday’s hot ideas.

The lesson to be learned from guys like Barry Diller, Steve Ballmer, Jerry Yang, Larry Page, and Sergey Brin is that you can make a lot of money from one idea but sooner or later someone else will come up with another idea that leaves you in the past. The next set of search billionaires may or may not come from the ranks of current search engine employees, but I’m not betting on that. No matter how much Google and its competitors try to stack the deck in their favor the odds are against their long-term relevance to the world of search.

That’s the price you pay for being intractible in a world of constantly changing technology. In search engine optimization you have to be flexible and willing to try new things, even to abandon all your old ideas. That’s the SEO Method: Experiment. Evaluate. Adjust. It works regardless of whether you’re trying to get into the index or to be the index.