I finally bit the bullet and read Nate Silver’s Bayesian prediction book that’s made so much press from predicting election outcomes and such. Nate’s that guy who uncannily predicted the outcome of EVERY electoral college vote in the 2012 presidential election. Okay, I’m now on the Nate bandwagon in regards to my field of SEO… well, sort of. But as usual, with all great books and great ideas, it just reinforces something we already intuitively know—or at least suspect…
The world is way more complex than can be pigeon-holed down into a simple computer model. On a lot of unpredictable subject-matter, the model can only be directional in nature and must be open to constant new input, which has been “pre-filtered” or “pre-considered” by a well-informed, mostly emotionless HUMAN! That’s a big part of Nate’s message: a bunch of stuff is extremely predictable, and a bunch is not… and those pontificating assholes should be a little more honest about it.
However, in those cases where things ARE predictable, you have to zero-in on the real story with a rinse-and-repeat process that uses new input as it becomes available, understanding the fact that even this new input is in fact subjective—yet, it’s striving to be impartially and emotionlessly objective as possible. Get it? If you do, you’re in some tiny but influential company.
Stated another way, under Nate’s model, there are TWO SIDES of Bayesian probability… especially when it comes to SEO: precise content recommendations that can probably and pretty accurately be done in analytics—versus the much more fuzzy-stuff regarding what direction the field of SEO is headed as a whole. Can you use Bayesian probably to answer the question: “Is SEO dead?” In this article, I address the second type: what does probability say about what’s happening with the field of SEO as a whole? Is it dead? Should I switch fields?
If this sounds a bit wacky to you, then you’re right along there with everyone else who had been ignoring the Bayesian method of probability prediction, due to its abandoning of absolute measurable factors, replaced by an “eyeballed” percentage provided by a keen observer. When this happens, predictions improve over time, as updated data (truer scenarios) are fed into the system. Wacky indeed! In fact, Bayesian logic was lost to history for over 200 years since articulated by mathematician and theologian Thomas Bayes in the 1700′s. Frankly, it sounds a bit like fufu math that no one should buy into… oh, except that it works. Why?
Things that are as subjective in nature and subject-to-human-error as the Bayesian process just doesn’t sit well with pure math and statistician-type folks. And it is true that accuracy of predictions even with Bayesian methods DROPS if the human involved is excessively ruled by emotions, or excessively vested in certain outcomes. Biases amplify. But another type of person feeding the data in for predictive models can get uncannily accurate results, as Nate himself proved with baseball stats and elections.
The best users of Bayesian logic make continual observations and input probability adjustments that are dispassionate, multi-disciplinary, well-informed. In short, they couldn’t give a shit about the actual outcome. They are ideal dispassionate observers—which rules-out just about every TV pundit, or career-minded “expert” peddling their particular view. Folks who can achieve that oddly accurate predictive behavior are strange beasts indeed—neither fish nor fowl—and probably not IN the industry they’re making predictions about.
Such successful predictive models CAN NOT be applied with confidence to just any situation. The subject-matter for predictions have to follow a fairly consistent set of rules. Hence, why it worked so well with baseball and electoral college elections—the places where Nate made his reputation. His new book, The Signal and the Noise, is throwing the world into a tizzy looking for where such uncannily accurate, if not subjectively derived, predictive models can be applied. The weather? Stock markets? Personal growth and career paths?
Of course, I look at the world of search engine optimization (SEO) as my current career, in dire need of some predictive science. Will Google+ take off or quietly ascend to dominance, swapping out the infinitely corruptible Link Graph with the easier to quality-check Authority Graph? When will the pendulum swing in that direction? Will Google hold the keys to the Authority Kingdom, or are there openings for other players to transform the popular concept of “search”? What role does a Googlebot web-crawl based search still play in light of the purely crowd-sourced Facebook Graph Search? How many people still like the anonymous but massive and massively-corruptible long-tail versus the the cleaned-up, curated, and much more mainstream content that populates Facebook as shared stories in the prattle-stream?
Precisely how this “great game” of search and online habits is going to play out is… simply… unpredictable. There are too many variables. The success or failure of these endeavors is determined by implementation details that we are not yet privy too. It’s much more like the stock-market, a much more unpredictable system, than like baseball statistics or electoral vote outcomes. The input parameters are unknowable—and even if known, with insufficient details to make an accurate percentage outcome prediction. Will Amazon start giving Kindles for free to Prime customers to predispose their online experience? Will it be Android-based versus Tizen or Ubuntu? What’s the formula for dark horses, long-shots, and unknown territory?
Oh, I project a voice and an opinion that sounds a great deal like accurate predictive capabilities, but I would argue that they’ve been delivered to me by captain obvious. I just articulate what’s going on better than some folks. Obviously, the Internet was a mistake as far as big publicly traded companies are concerned. Its existence goes AGAINST the bottom-line of profits and revenue, because with better information in peoples’ hands comes more informed purchasing decisions and more opportunity for newcomer competitors to upset the apple cart. Actually superior products and services get an edge over the fatcat incumbents. Fatcat incumbents want to correct that… correct the Internet.
To that end, they desire to predispose your very online experience through the device. He or she who equips you with your primary preferred device to access online services has EXTREME ability to predispose what that experience is, and what products and services to which you are ultimately exposed. Think cheap tablets. There is already a battle over the pricing of the 7-inch form-factor. Kindles set it at $200. Google Nexus 7 upped the ante of what you get for $200. India’s Akash gives you SOMETHING for $50. Acer’s TRYING to make Nexus 7-quality available for $99. And that’s all just THIS YEAR.
The Internet was originally designed to survive nuclear blasts, enabling “packets” of communication to “route” themselves around the disruption. On a big network, there are infinite paths for the packages to dynamically adapt themselves to follow to get you the info you requested. It’s quite a robust system, great for freedom, democracy, and the informed consumer—but not so great to the incumbents being knocked off their high horses. Entire empires fell from this new dynamic, and new sexier ones sprung up to fill their space.
Ah yes, the four horsemen… or is it five? Or six? Is Samsung going to become a diverse products and services company? Can Microsoft regain its footing now that the world has collectively knocked it wobbling for delivering too little too late in the shadow of Apple? Apple, Amazon, Google, Facebook… and maybe Samsung. And maybe Microsoft. Oh, and does IBM still factor in at all? They’re doing some nifty stuff with search and big data, as many of us saw Watson crushing humans on Jeopardy. And is Facebook really that savvy, with no likely plans for hardware?
It all started with Apple turning worldwide assumptions upside down by proving hardware is indeed important. Hardware is where the personal relationship comes into play—and personal relationships are everything. People LOVE their iPhones. Perfectly milled cases that feel more like jewels in your hands and pockets matter. But not only that, Apple’s odd draconian rule over its iTunes ecosystem, operating system, hardware, and even right down to the programming language for apps (Objective C) and adherence to rules regarding API-usage for performance… well, they pulled off a piece of social engineering artwork the likes of which… well… perhaps probably the world has NEVER seen before. Except maybe Commodore.
As many of my long-time readers know, my heart belongs to this long-dead loser—a loser who mastered the total vertical integration game to deliver superior hardware at a cheaper price to a broader audience—long before anyone even understood that was the game. They got control of the world’s $25 processors when everyone else had to use $300 processors, making Commodore able to deliver a complete $500 box when everyone else’s were $3000. But Commodore screwed up a lot of the little implementation details that Apple mastered.
And so now we live in a different, much more educated world. Hardware matters. Superior performing hardware at ever-cheaper prices matter. There used to be this eternal “Price, Quality, Speed – pick any two” game that is no longer true. As much as people like to bust on Apple, the iPad for example delivered price, quality and speed at about a $500 price point when the equivalent netbooks on the market, the Asus Eee 700 model from the prior year for example, delivered only price and… Hmmm. What was the second thing? Yep, I owned it.
And so obviously, the world is now scrambling on how to compete with Apple on all three fronts: price, quality and speed. Now speed doesn’t refer to computer-speed like MHz and GHz CPU processor speed measurements. In this context, speed refers to the manufacturer’s ability to innovate and keep bringing newer and better products to market at forever lower prices, with forever better specs and battery-life. But who? Amazon with Kindle Fire? Close. Google with Nexus 7? Yep, but it wasn’t Google. It was the same good ol’ Asus who brought us netbooks. Oh, and NVidia who had to reinvent themselves or die.
Not to digress too much in this article that’s ultimately about SEO and the science of prediction. But, NVidia seeing as how Intel just made them more-or-less obsolete by including “good enough” [for gaming] graphics co-processors directly into the current Ivy Bridge generation of Intel chips designed to power the Macbook-air competitive line of Ultrabooks (the design-specification provided by Intel as a gift to the industry), NVidia responded with the high-tech version of a bitch-slap. They made their own gift to the industry: how to design high quality cheap TABLET computers around the ARM processor—a technology competitive to Intel, which powers all our Android and iPhone phones. It’s called the NVidia Kai platform, based on the NVidia designed Tegra 3 soc (system on a chip). And no one’s heard of it… except as the Nexus 7, of course… built by Asus… marketed by Google.
Oh, what a tangled web! The unholy alliance of NVidia, Asus and Google is EXACTLY analogous to the unholy alliance of the 80′s between Acorn, VLSI and Apple that gave rise to the ARM processor technology. ARM was originally designed by a desperate UK-based competitor of Commodore’s, built by one of the best chip fabricators of the day, and funded at least in part by a US-side interested third party that ended up using the thing in the Newton. Here we are thirty years later, and one of the dark-horse unholy alliances of tech passes the baton to another new unholy alliance of tech… this time, the ferocious competitor not Commodore… but Apple!
What the hell does any of this have to do with SEO and the Internet, anyway? Well… the game is fairly static so long as the Internet as we know it is accessed through devices that cost a lot, and get infrequently revised and replaced. There is little opportunity for major seismic shifts. The network is generically the network, which is generally the Internet. Access to it is generically hardware, which is generally your desktop Windows PC running Internet Explorer.
Get it? The game isn’t static anymore. Because of ever-cheaper hardware with ever-faster release cycles, the network is NOT NECESSARILY the generic Internet, and your device to access it is ALMOST CERTAINLY no longer your desktop PC. This creates an opening for seismic game-changing shifts. This is where the four, five, six horesmen play. But it’s a checklist of items that must be wholly owned—or at least, mastered in order for one to rise to dominance. It’s why Apple owns hardware and digital goods. It’s why Amazon owns datacenters and hardgoods fulfillment houses. It’s why Google owns information and bought Android and Motorola.
The checklist is: you need the hardware portal to access services. It’s gotta be cheap enough to frequently improve and drop new models into your hands every 2 years or so. There’s gotta be a digital economy where everything that CAN be bought or sold in digital form (music, movies, books, etc.) is sold. You need the massive cloud-like datacenter infrastructure to support this—servers on top of servers on top of servers. The server infrastructure needs to be seamlessly upgradable to benefit from Moore’s Law.
The cheap high-quality tablet-like device that keeps getting dropped into your hands needs to have as many pieces as possible provided by YOUR OWN company or COMPLETELY free and open source (FOSS) software (if there is such a thing), so that other companies do not have choke-hold on you. That means Amazon, Facebook, Samsung, and obviously Apple, must get Android off their product—much more realistic than you may think, since Android is a ziggurat of shortcuts—Java on top of Linux with a bunch of custom drivers—that makes Android apps uniquely portable to Tizen and Ubuntu Mobile. Remember, all Ubuntu needs of Android is those delicious hardware-specific drivers for cameras, graphics optimization and such.
Ahhhhh, so if you want to own everything about the tablet that accesses online services, it’s clear that Tizen and Ubuntu are paths off of Android. But can you EVER wean the world off the Internet, or somehow limit its access or exposure, so you can always peddle your own products (or your advertiser’s products) first, before the experience provided by some other company who weaseled their way onto your hardware via impossible-to-stop HTML5 Web apps? Amazon just did this to Apple with an Amazon web app “site” iTunes store competitor. It violates Apple’s terms of service, but what’s Apple going to do when music you bought over the Web can stream over the Web? The Internet itself is a difficult-to-control factor that must be mitigated. WTF did I just say? Jeff Bezos has some kung fu that rivals the late great Steve Jobs. We just haven’t seen Jeff do his thing yet… and I think he might get a hand from another kung fu master, Mark Shuttleworth.
Apple uncloaked its brilliant ecosystem maneuver with the iTunes economy. It’s a play that’s equally as brilliant as what it did to break its dependence on PowerPC processors, and is once again doing to break its dependence on Intel CPUs. The dependency on the Internet is broken simply by having a superior, curated experience in the iTunes and App Store environment that gives you everything you feel you might miss about the Net: search, crowdsourced reviews, cool software, your news and other stuff VIA those apps… oh and what the Net CAN’T provide you: a sense of security and a seal-of-approval by the ghost of Uncle Steve.
Ahh, iTunes. This is where we mention one of the checklist items the N-horsemen must also provide: a friction-free payment method. Apple tied it to your phone-bill and credit card, but they DID pull it off. You can make a $1 purchase at a time without hardly thinking about it. At worst, you have to provide your Apple .mac .me .icloud .whatever login credentials, and BAM! $1 billed against your credit card. Amazon did something similar, and took it to a new level with Prime and subscription services that just keeps you buying. Google struggled with Checkout… Google Wallet… Google Play… whatever they’re calling it. What will Facebook do? Is Samsung even in this game? It’s another place Microsoft did too little too late. Apple’s brilliant move was to drive the process with music, and then later through curated apps.
Hopefully, I’m drawing a picture for you of giants each trying to drop different hardware into your hands and get you addicted to their particular mix of services and experience. Not all of the horsemen can easily reproduce the killer-app checklist item of their competitors. For extremely large-scale cloud services Apple had to source it from the Amazon and Microsoft clouds (Microsoft Azure). Google is struggling to outsource all the Amazon-like hardgoods fulfillment to the world, because reproducing a $10-billion Hoover Dam-scale pick-and-pack operation ain’t easy. Maybe Wal-Mart could do it. But Google? It seems antithetical to their information-centric existence. But Google has to do SOMETHING like it to compete with Amazon. A horseman’s ability to outsource these checklist items is uncertain. Just ask Apple in regards to maps.
Are maps a checklist item? Nope, but the integration of local product or service offerings WITH maps is. Or in other words, replacing the function of the old-fashioned Yellow Pages and White Pages phone-book IS a checklist item of this new breed of megalifestyle-provider companies. Huh? Okay, a definition: megalifestyle-provider company IS a horseman. It’s something akin to what GE was when it owned NBC, both forming all the products of life, and then programming the public to use those products. The Net put the kibosh on GE/NBC because it needed efficient mass media to work—but that only left a vacuum that the horsemen are trying to fill.
So far on the checklist, we’ve got cheap slick hardware, an autonomous OS, a friction-less payment system, a digital goods marketplace, hard goods fulfillment capability, a massive cloud infrastructure, and local data to fuel maps and local search. We had better throw in cloud-synced Office software suite. But is that all of the checklist items?
I hinted at making the Internet-proper unnecessary, but only an AT&T, Verizon, Comcast or major telecom could do that, and government anti-monopoly laws would clamp down on it—so, as awesome as those products would be, it’s unlikely. Remember how early Kindles connected to Sprint with users hardly thinking about it? Could someone really pull off an alternative to the Internet proper? Telecom independence is the elusive unattainable checklist item. As his biography reveals, it’s a holy grail that even Steve Jobs chased in his early vision for the iPhone. A universal truth that even the reality distortion field can’t change: highly reliable and available bandwidth takes towers and a whole bunch of satellite launching and tearing up the ground to lay fiber that… well, not even Apple can just bully their way into. Telecoms are special, and are kinda immune to getting rolled into a horseman ’cause of consumer and government anti-trust watchdogs. Content is a bit easier to wrap-into the checklist, and that’ll be the subject of another article.
How about an advertising network? Ad revenue to offset hardware cost? A way for people outside the new ecosystems to buy their way in? Certainly! Who would leave such vast quantities of money on the table? Display ads on web pages and in apps, and ad-insertion on search. Just imagine if Amazon put out an Ubuntu-powered Kindle whose search hit Amazon first. Far-fetched? The current released version of Ubuntu 12.04 DOES PRECISELY THAT on the desktop in the desktop’s easily-accessed built-in search feature called Dash. And now Canonical announced Ubuntu Mobile that can be installed on any Android-compatible phone. Let that sink in for a moment.
So, who can predict? Everything I’ve done is mere observations, and guess-work based on all the facts around us. Dots are popping up all the time that need connecting. But even the picture drawn by the connected dots may never come to be. Could Amazon REALLY EVER put out a non-Android Kindle Tablet? And if they did, could they give it to all Prime customers for free? And if they could do that, could it be a good enough experience to make you NOT MISS Android, or your iPhone or Google or the Internet in general? And that’s JUST ONE scenario that could play out in parallel with a half-dozen other game-changing seismic shifting plays by the megalifestyle-providers that some call the four horsemen.
This is perfect ground for prediction. So, using Nate Silver’s prediction methods here are very appealing. But what variables would you fill into the now famous Bayesian equation? It is unclear to me. There are too many variables of unknowable values. It is much different than baseball statistics or electoral votes. Yet, there is something vaguely predictable through a certain type of obviousness that arises from being a student of history and fan of technology.
Human nature and the free market being what they are, and certain things having no other possible outcome (better, cheaper, faster)—certain patterns WILL rise again. And those patterns are the re-assertion of corporate influence over today’s free-for-all Internet via the embrace-and-displace strategy—often used by Microsoft during its rise. Specifically, embrace-and-displace will take the form of building neat little devices that Free Software advocate Richard Matthew Stallman (rms) would describe as “beautiful prisons”.
As an iPhone customer, literally from day-1, I LOVE my beautiful prison. I appreciate Apple protecting the sanctity of my phone against malware criminals phishing for me through corrupt apps and even more corruptible operating systems. Sandboxing apps, granular permissions, and accessing phone features through strictly controlled API-calls are concepts I endorse—ON MY PHONE! My thinking changes dramatically when it comes to how I program and write server apps of my own.
When you are a programmer, there should be absolutely no limits imposed on you by the platform that are not limits that you chose in exchange for other advantages the platform provides. Fully explaining that statement will be the subject-matter for another article. But in-short, there are places for beautiful prisons. Mikey likey his beautiful prisons. AND there are places for totally free and open source environments where you can transcend the abilities of the normal public.
And THAT is precisely what I do in inventing SEO systems like HitTail.com, which is enjoying a sort of renaissance, given its motivated new owner Rob Walling and the strange SEO days brought about by the Google Panda and Penguin algorithm tweaks and war against spammers. The long tail really does work well, and I know this through my continued hittailing, using the very product I invented to help predict the future. I did something similar with 360iTiger.com, but that has not reached maturation yet. Stay tuned on that front, but it’s ANOTHER great source of the as-objective-as-possible, but still-quite-subjective data that feeds the Bayesian theorem.
I can’t really predict the future in terms of what’s going to happen to the field of SEO, and natural search itself—but I’m going to try. I haven’t even talked about Siri, Google Now, and the attempt to turn old-fashioned crawl-based Web search into a sort of backfill or overflow of search—but a brilliant coworker of mine, Boris Zilberman, has done precisely this in this article just published over on Search Engine Journal—so go read that also.
Bayesian probability likely indicates cheap awesome hardware and its frequent upgrade cycle as the likeliest driver of change. What are the odds that Android can be knocked off these platforms? What’s the percentage probability? That would reduce the impact of Google AdWords revenue on the future of mobile, and make yet another decision-branch. What would that be? Can an adequate experience be provided on such a theoretical mobile platform WITHOUT Google? Yes/No? Percentage likelihood? Is the ecosystem sufficiently complete and satisfying to isolate the tribe on that platform from the broader Internet-accessing world? Yes/No? Percentage likelihood? And so on. I will keep zero’ing in on the truth.
And how important is the Google that’s left standing? If successful with Android, Motorola Mobile, Google+ and being the owner of worldwide Author Authority, then Google remains very important for SEO, and the field of helping Authors and Agents with their Authority Rank becomes the new sweet spot in the field of SEO. If not, then it’s jockeying to provide the best data feeds into this strange heterogeneous mix of best-of-breed search experiences each horseman is going to try to knit together. Influence Wikipedia, get in Yelp!, etc.
Given all the search engine landscape stuff that I fleshed-out above, my decision is to act on what I know for sure so far to in order to assure the best future for myself, my family, and education for my 2 year-old daughter. I’m betting on BOTH being an Authority Authority (Google+ is not what you think), and on being a general API-using wizard—whatever you want to call that—look at 360iTiger.com and Levinux for a clue.