Mike Levin SEO

Future-proof your technology-skills with Linux, Python, vim & git... and me!

AgentRank is the future of SEO… someday.

by Mike Levin SEO & Datamaster, 09/12/2012

This post is a direct continuation of yesterday’s ramblings. I discussed the stream-and-tag model of organizing content for search versus the crawl-and-index model. In the stream, you have a whole bunch of content authors and publishers wanting their stuff to be discovered and perform well in search, so they tweet announce their content, Facebook share or Google+ post. Information consumers and curators step in and start tagging and associating their own identities with the content. Identities are not equal, because some Google+ profiles have ascended to authority positions on certain keyword topics.

While such a system initially serves to elevate the super-popular pop-culture silly meme content even higher, over time real domain authority experts will sort out and lend their credence to the less mainstream but still meritorious content. It’s going to be an iterative and long painful process, just like PageRank. And Google insists this will not replace the link graph and PageRank, but rather work an tandem—and I presume become merged with it.

Google insists that the crawl and the link graph and PageRank aren’t going away. And Google’s Web Quality Team, or whatever Matt Cutts group is being called today, is taking it up several notches informing people what’s wrong with their sites through webmaster tools and emails. Simultaneously, the algorithmic war against spam continues with Panda and Penguin. I rather suspect the link graph has entered a state much like email had in the early 2000’s: mostly spam, threatening to make the entire email system useless.

Now it’s not that spam is everywhere. It’s just that it is where it’s most profitable. Profiteers were saturating the Internet with hyper-optimized content where buyer intent was highest, gobbling up all that editorial search space thanks to their technical savvy. That’s the game Google created, and because there is a blurred line between building links and link manipulation, it is also a blurred line of hypocrisy Google must adopt in fighting spammers. It’s all about thresholds.

So aside from algorithm and threshold tweaks and action by the Web Spam Team via Webmaster Central, what has Google been doing? Well, they’ve been taking a page from the spammers themselves and monitoring that most profitable search result space with more ads. More screen real estate is clearly going over to ads when the queries indicate buyer intent. Google Shopping results when they appear in result-space (as opposed to the upper-right) has been the most glorious example of late.

Now, this is Google’s right. Google is just like any other ad-supported media—TV, magazines, etc. They have to strike a balance between the ads that pays for the free content and the content itself. On other media, that free content is articles in the case of magazines and TV shows in the case of TV. With Google, it’s “editorial search” often mislabeled natural or organic search. Presumably, gathering its data with a crawl and sorting it with objective and unbiased (by ads) makes it organic. But in my book, that makes it editorial.

Running these increased ads is a double-win for Google. It both raises revenues and diminishes the impact of spammers where buyer intent is highest. So you can almost visualize an attack on three fronts: ELIMINATION in which spam filters pruning out offensive sites out of results, PUSH DOWN in which ads push all but one or two results down off the screen on valuable terms, and VERIFICATION in which known expert authors (or “agents”) testify as to the merit of the content that remains.

Content is being sifted and sorted. The cesspool that is the Internet, by Google’s own admission and in great part by their own doing is undergoing a cleanup. Tons of cruft is going to be left over, but that too will wash away as throw-away domains are not renewed and spammers stop paying to host now ineffective content. All sites with a pulse and a mission in life will be overhauled with HTML5 and tied-to/buoyed-by verified authors. Gradually, the cesspool once again becomes merely a dirty outhouse.

Google’s formidable crawling infrastructure keeps chugging away, but now with a slightly demoted role of discovery and long-tail overflow results to researchers, curiosity seekers and all the other non-buying-intent uses for search. For a decade, this “librarian-like” research behavior got center stage, because the world was infatuated with how much there was to know, and mainstream media failed miserably in fixing the publics attention on their own competing media.

But that was an anomalous decade. People are starting to “get it” and there are new category-killer starting-points on the Internet than Google. For shopping, there’s Amazon. Compare Googles 1-in-5 search that can be monetized to Amazon’s 100%. Talk about buying intent! And Amazon has its own tablet, mobile apps, and pretty soon, its own phone! It had its own search engine with A9 but abandoned it. It really doesn’t need it. It pre-cherry-picked the best types of searches in the world: shopping. Oh yeah, and it’s got the datacenters and physical product fulfillment infrastructure to do almost anything. Fear the dark horse, Amazon.

Twitter and Facebook are of course category-killer starting points on the Internet too. But all that influence people are building up in profiles is hard to cash-in. Facebook is not organized to surface domain experts and give them clout in their field, so as to have dominion over their field. Twitter actually is structured a little more like that being open by default, but those influential profiles are effectively squandered by not… well, by not being part of Google. Firehose deals for access to the Tweet Stream such as Microsoft Bing has will just never allow default search to make as much fundamental baked-into-the-DNA of search that owning Twitter could.

Yes, almost 20 years into the Web as popularized by Netscape, and 15 years into the world reshaped by Google, a fundamental shift has arrived. Social, mobile (and local, by extension) are the drivers of this shift. There is a new link-graph that works takes into account real people, real identities, and real expertise built-up over time. But 2 of the 3 social link graphs on the planet are in prison, as far as generic search is concerned. The third social link graph has been described as a ghost town.

Experiments with Google+ show that individual posts are not necessarily much of a relevancy factor yet in controlling search. What’s further, there is no way to know how far Google has gone in implementing this new AgentRank scheme that leverages author profiles. We’ve got some clues in that the thumbnail pictures of authors are indeed appearing next to signed content that uses rel=author. But that’s just an initial baby-step. That’s tea leaves that need to be read.

If AgentRank turns out to be anything like PageRank, its going to be on an exponential scale. That is, the most influential people are not going to be 10x more influential than the least. They are going to be thousands or millions of times more influential. It’s like the Richter Scale to measure the magnitude of earthquakes. Each number on the scale is over 30x the energy of the previous number, so that the whole range of known earthquake intensities can be laid out on a scale of 1-to-10, from barely noticeable to armageddon.

For people to raise to enough authority to have real result-controlling influence, they’re going to have to work their way up that scale, which will be its own competitive game. That means, you’re going to have to be included in circles, and those circles have to be NAMED with the keywords you want to be influential on. For early examples, you can check how Danny Sullivan will be influential on SEO by just searching on the term SEO and seeing his profile plopped up into the upper-right. He’s in over 1.5 million Circles—probably all named with SEO in it!

It’s a less smoking gun example, but google Marketing Guru, and notice Seth’s Author-verfied Typepad blog near the top of results. Now, Seth never posted even a single thing in Google+, but he did at some point yield to Google’s pestering to upgrade your gmail account to a Google+ profile, and that opened the floodgate to people putting him in circles. He went the additional step of going through the Author verification process, and that sealed the deal. Now, I can’t say with certainty that this is why he positions so well on Marketing Guru, but with being in over 96K circles (a fact displayed right in the search results), named something like Marketing Guru, it’s a pretty good hypothesis.

AgentRank even becomes more compelling when you realize Google chose the term AgentRank instead of AuthorRank to address the concern by marketers that their companies are not people. Companies are virtual entities that can live longer than people and should be able to carry over the assets beyond the death of owners and officers. And so, Agents is the proxy for people when its a company. Google+ already has company profiles, and just started divvying out vanity URLs. They are extremely slow to move, but the writing is on the wall. Companies will have presences online just like people, appearing in default search with logos, authoring content, +1’ing content, and generally acting like a person online. This accommodates for some of what will be lost as PageRank goes passe.

How are the other companies going to respond? Is there anything even remotely like Author Verification and AgentRank from the other companies? And is identity really even the killer relevancy criteria for search? Well, as far behind Facebook and Twitter Google might be on accumulating social data, Google is ahead of them in building search infrastructure. They’ve also got the brand loyalty, which although I view it as fragile due to no customer lock-in, it is still ten years of difficult-to-break habits weighing in Google’s favor. The question is whether that is enough to survive the waiting game as social data accumulates and AgentRank authority authors begin to ascend—before disruption from some other direction occurs.

Well, of course many people would answer it is Facebook with its fan pages and followers. I would argue that Facebook actually requires TOO MUCH buy-in to participate. I’m on Google all the time doing real research, and if I make it to Facebook once or twice a week, I’m lucky. There’s just something about how Facebook is built that keeps it from being a play at reaching out to the world. It’s not a good inbound or permission marketing vehicle—probably the source of a lot of its recent troubles. Facebook really is like the automated family newsletter, and how welcome are ads there? They are very much the equivalent of commercials on TV. They missed that magical serve-on-inquiry moment that Google AdWords enjoys.

So, where else could that disruption come from? Well, some folks are like Blekko and DuckDuckGo banking on early adopters becoming disenfranchised with Google selling out to advertisers, and looking to switch to a simpler search engine. Others think that maybe Microsoft’s flailing around for just the right magic deal and good implementation could stumble onto it and transform Bing into that social search engine. This would be accelerated if Apple switched their default search over to Bing… or perhaps, Yahoo if they couldn’t swallow that.

Did I mention Apple? Well, of course Siri was an attempt—or perhaps a modest first step—towards search disruption and an attack on the crawl-and-index model that’s the last thing in the world anyone wants to reproduce. In such an API-first model, editorial search is more honestly demoted to the unprofitable overflow backfill that it is. All the API data providers such as Yelp that need to come before the overflow are companies that now need to cut deals with Apple. It’s a much more curated and deliberate search experience than Google has traditionally been, and it’s on 30% of mobile phones.

Google followed suit by releasing Google Now that has arguably better voice recognition and results in the latest Android OS, Jellybean. Also, the Android platform—as fragmented and difficult-to-upgrade as it may be—has really taken mobile by storm, somewhat fortifying Google against a death-blow attack on search in mobile. Those people who I keep reminding you are users (not customers) are incrementally becoming more of customers as they pay for smartphones and tablets that run Android.

Some old wisdom is being overturned here. It goes that the incumbent becomes old, lazy and tired as some young nimble competitor realizes something new is possible, goes and changes the world, and the incumbent can’t adapt. Incumbents can adapt to some degree, as we saw as the world switched from Mainframe computers to PCs, and IBM managed to make the transition, but you have to continuously change your trick. IBM alternates between being a hardware and a service company to survive.

Some other old wisdom that founders are more in the way than an asset is also wrong—the wisdom going that founders are a whacky bunch good at building, but terrible at long-term strong business growth. A first generation of PC wizards like Bill Gates deflated that rumor, and the seminally technically adept artist Bill Gates knocked that theory on its ass. But you have to re-invent, as Apple did with music. If you’re a one-trick pony, like Michael Dell, you’re not going to be able to counter this wisdom. Larry Page got back his president status from the temporary grown-up placeholder Eric Schmidt. And Mark Zuckerberg never gave up that control in the first place. Oh, and there’s Jeff Bezos, of course trotting along a brilliant path less traveled.

The clash-of-the-titans story of our times is over: Apple Vs. PC. Steve Jobs Vs. Bill Gates. Apple won. Microsoft is on its way to being relegated to a mere IBM. Now that’s not to say it’s going to fail, or even start losing profits. But it no longer dominates the game and defines our world. Microsoft becoming a mere IBM is still a very strong position. Apple won, because they realized you have to have a balanced formula and love-worthy products to win peoples hearts and minds—and use that advantage as a beachhead in winning more ground.

The new clash of the titans story is Apple Vs. Google—with Amazon sitting by the side taking notes and practicing. Google’s the heavyweight one-punch-knockout newcomer but with a glass jaw, because it’s got no customer lock-in. Apple’s the old champion who knows every trick in the book and has long ago practiced away its vulnerabilities, but doesn’t pack quite a wallop in the one-punch-knockout search department.

And so Google fights to become more Apple-like in releasing the Nexus 7 and buying Motorola Mobile. They even fight to become more Amazon-like in refining Google Shopping and Google Play. Google is working to fix its glass jaw and make some of its $40 billion in annual revenues come from something other than AdWords advertising in vulnerable search. But Google’s never going to build our a physical product fulfillment warehouse operation like Amazon, and Amazon and Apple are probably not going down the crawler-based-search route. And none of them own enough social date to change the search game tomorrow.

Where that leaves us is in a strange sort of limbo where things get really interesting and exciting for the consumer. Devices get cheaper. Services get better. But the final dots fail to be connected for years. Google desperately wants to move into the future, but has to take a measure-ten-times, cut once approach because of the finickiness of users, and the speed with which you can lose goodwill assets. So we see a sort of ebb and flow on Google’s results page, and increasing number of things they can do with social data before enough critical mass accumulates to have an impact on positions.