Rethinking a Career in SEO
by Mike LevinSunday, February 12, 2023
Where to begin? Well, at the beginning naturally. And then we’ll work our way all the way through to the end. Things are different now. Google was an almost 2-decade anomaly. AltaVista from Digital Equipment Corporation (DEC) in 1995 could have been Google had they had something like PageRank to improve relevancy, but they didn’t and Google did. Next probably a few years later 1n 1998 when GoTo.com (later, Overture.com) added pay-per-click results blended into their Inktomi results teaching the world how to truly monetize search results, it was another time when Google could have been walloped, but they were quick to wrap in this idea from Bill Joy of IdeaLab into Google in the form of AdWords. Google didn’t go right to the PPC model. GoTo.com showed Google the way. And now we’ve have the better part of a 2-decade run of not even Yahoo, Microsoft and the rest of the world being able to displace Google’s search dominance.
Until today. Now I regularly go to ChatGPT before I go to Google, namely to do my research for articles exactly like this. We are in the early days where the OpenAI-powered ChatGPT can’t cite sources and I have to back-up my information gleaned from ChatGPT sessions with follow-on Googling. But ChatGPT now comes first to help flesh out overarching narratives without having to deal with the sifting through links and ads. Links and ads and ads and links. All Google has done was replace the old telephone White Pages and Yellow Pages with a new sort of directory pathology. Directories are pathological. They don’t last because they are magnets for competition and corruption. We’ve seen it happen with the Yahoo human-edited directory. We’ve seen it happen with The Open Directory Project (DMOZ). And it’s probably happened with Wikipedia, but they manage to keep it less visible, but Wikipedia is also a corruptible directory.
And so “real intelligence” to the rescue! There are various ways to go about it and ChatGPT took one route. It’s route allows it to lie when pressed for facts, has no opinions, and rarely cites sources. That’s corruptible in its own way and the current honeymoon we’re on with ChatGPT won’t last long. But that’s okay because the infinite resources of Microsoft to the rescue! The data is only from 2021? No problem! We’ll update it with real-time search results from Bing. The data doesn’t cite sources! No problem! We’ll tweak the model we’re using and layer in a few UX-hacks so that it can. And we’ll announce it, show it, start taking $20/mo for a premium service right now, and we’ll use the bully pulpit of controlling the Windows operating system and default browser (Edge) to consistently slip this new ChatBot experience where where you would have normally encountered Google. Cut them off at the pass and coral them back in. Expect a browser default search reset real soon.
So a giant reset button is about to be reset and an industry recalibration is about to occur. What’s important and what’s not important in the online landscape will subtly change. The insane online development of infinite new “properties” and “real-estate”, which while technically is forever easier to perform, is also technically less effective and therefore less worth doing. Content production techniques that don’t add value and only do that whole meta remixing routine are now being recognized as such and ignored.
The corollary is unfortunately not also occurring. Tuning-down the spam without tuning-up the wholly unique (the truly E-E-A-T) but undiscovered content of the net, because it’s inbound links aren’t over some threshold or whatever, will generally continue to be wholly unique and undiscovered. The ability to ferret out this sort of content to serve up in results (be they AI-answers or lists of links) will be one of the truly differentiating points of next gen search.
The direction next gen search is going is actually not that great for the little guy, because it doesn’t surface new stuff. It favors incumbents more than ever. And among the incumbents, it seems to be surfacing things that work like news organizations and fit into the Google Discover model. It’s the old days of RSS News readers like Google Reader, but Google doing the curation for you. I feel aspects of the YouTube algorithm that drives up the ad-clicking slipping into Google Discover for sure, and perhaps even default web search. It’s a polished and pandering publisher’s world at the moment, and only the “made ones” at that.
The old PageRank system is still lurking there under the surface. It has a dampening effect literally built-in so that things don’t reshuffle too quickly between crawl/index cycles. A lot of old incumbents are being pinned-up in results due to this, but just like the frog getting slowly cooked, the influence of PageRank gets dialed down while the influence of the other systems, no longer layered-in and like the old Universal Results, Knowledge Graph and various carousel systems get replaced by complete alternative