For the first time in over two decades, the internet’s most ubiquitous start page (Google.com, of course) could be flirting with an extinction level event... and one that’s partially of its own making.
And as unthinkable as it sounds today, historically, this is normally how industrial empires fall - not by being outwitted by competitors, or from bad management - but rather by failing to register a rising threat in their blindspot. Or otherwise failing to take it seriously.
This is what Google faces currently; it’s known as the Innovator’s Dilemma. Namely, this is when a successful company has far more incentive to maintain the status quo versus pursuing experimental markets or riskier innovations - especially if it potentially undermines their core business model.
Some examples:
In each of these cases, and many others across history, these failed companies usually made the “most logical choice” at the time. Maintaining the status quo was the surest bet, and any competent manager knows that it’s better to stack up base-hits vs. going for the all-or-nothing home run.
As Google stares down the barrel and faces its own Kodak Moment, there is a legitimate chance that we’re now seeing the final days of its core product: The Search Box.
Let’s examine what’s happening - and what this could mean for publishers, advertisers, and the internet as a whole...
PART 1:
While Generative AI has technically been in development since the early 2000’s, it’s only in the last few years that this technology has progressed to a level where it’s accessible to everyday users.
Ironically, Google has been instrumental in advancing the GenAI space through its programs like Google Brain, DeepMind, TensorFlow, and others. So while not solely responsible, they certainly played a part in building their own proverbial “Frankenstein”.
In particular, things really started to accelerate when OpenAI released GPT (short for Generative Pre-trained Transformer), which is the engine that powered the first wave of GenAI apps like Jasper & Copy.AI. And today, it’s used by thousands of similar startups.
But it wasn’t until their release of ChatGPT in November that the world had its collective aha moment, and we all had our first real “conversation” with the intelligence behind the screen... immediately followed by an existential realization.
“Oh my God... this is going to change everything.”
In just 2 months after its launch, ChatGPT saw its 100 Millionth user, making it the fastest-growing app, ever.
And what is ChatGPT, in simple terms?
It’s a text field and a submit button. Just like Google Search.
Except, as we all know, pressing the button with ChatGPT can do so much more than crawl the web.
Along with about a dozen other disruptive use-cases, it was immediately clear that ChatGPT was going to be an actual competitor to Google - and for many types of queries it’s already a vastly superior alternative.
Across Twitter, and even among friends who don’t work in tech or marketing, it’s now common to hear people say that they’re already using ChatGPT for half of their “searches”. And given that ChatGPT sees approx 1.7 Billion users a month (SimilarWeb), this anecdote seems to have some merit.
But it was Microsoft’s investment into OpenAI, followed immediately by its subsequent release of “New Bing” (powered by ChatGPT) that I think threw down the gauntlet and forced Google to rapidly unveil their own AI experiences.
Namely, Google Bard, and shortly thereafter, SGE (Search Generative Experience).
If the keynote for Bard was any indication, Google was hardly ready to go prime-time with any of these experiences, and my guess is they’ve been dreading going down this path, regardless of the timeline.
Unlike Bing - which has virtually no real market share, and therefore nothing to lose - Google truly has nothing to gain in this situation. All of these innovations are at best, defensive. And at worst, they’re actively destructive.
Why?
The vast majority of Google’s revenue comes from advertising. Specifically, 60% of its global revenue is just from search ads.
So, what happens as more & more users want to “talk” - or more accurately, interact with the world’s most tireless virtual researcher - rather than just being routed to a list of relevant websites?
We don’t know yet. Google obviously finds itself cornered by the inevitable shift from “Ask Jeeves” to “Jarvis”, and it’s clear that Bard & SGE are being released way before they’ve had time to polish and refine the experience.
This is also an area where Google’s size - normally an asset - suddenly becomes a giant liability. If they pivot too quickly, they risk sacrificing their core business model (PPC ads), which currently work so well based on the necessity of jumping-off to learn more.
There are also implications for its organic ecosystem: If they pivot too hard, they’ll disincentivize the publisher ecosystem (and its AdSense network). But if they don’t pivot enough, they’ll lose Search market share to Bing & others.
As marketers, what does it look like if Google stagnates and loses market share?
And similarly, what does it look like if Google leans in hard, and opts to sacrifice its short term revenues - along with its publisher’s share of traffic, and its advertisers’ EPCs?
Let’s examine some potential outcomes...
PART 2:
Content publishers have long had a strained relationship with Google.
It’s meant to be symbiotic; Google unofficially “agrees” to provide visibility for deserving content, in exchange for adding it to its index. In turn, publishers “agree” not to try and game the system, or otherwise serve up a poor user experience.
Obviously, it’s not a perfect system, but for the most part, this whale & remora relationship has largely worked out well for both parties.
And through its ad network (AdSense), Google is also able to capture value through much of this ecosystem even after its users click away from the SERPs; there are nearly 60 million website partners running Adsense.
But if Google leans in hard on SGE, this is going to change what it means to “rank”, users will have less reason to jump off to 3rd party content, and the amount of meaningful page-1 real estate will shrink tremendously.
As their own screenshots depict, this process will basically expand Google’s current Featured Snippets in the top-fold, give users a more detailed (generated) answer, and let them “chat” to clarify certain items.
As you can see, SGE takes up nearly all of the top-fold, and only links out to its “source material” on the top-right, as well as underneath the chat box area.
From this current format, we assume the following:
In addition to some wildcards like changing search behaviors, all of this ultimately translates into less traffic to publishers, period. Along with a complete re-think of what SEO means going forward.
Will the future of SEO be more about feeding “seed data” to the various LLM’s (SGE, GPT, etc) in hopes of being cited? Does that mean that the way to reach humans is to essentially write for robots?
Time will tell.
Now, another possibility is that Google only partially leans into SGE, and only deploys it conditionally for certain searches, as with Featured Snippets currently.
That’s a better outcome for publishers initially - but it also puts Google at risk of losing market share to competitors like Bing, as well as accidental competitors like ChatGPT, which would then, ultimately result in less traffic for publishers regardless.
Google has to strike a delicate balance between upholding market share while also continuing to provide incentives to its publishers - on whom it relies for timely, relevant content in the first place.
Already, the impacts of of LLMs like ChatGPT are becoming scarily apparent, as researchers hold up StackOverflow (and its rapid decline in human-created content) as a cautionary tale; their assessment is that LLMs are a threat to human data creation - which then becomes a threat to LLMs.
Last, while it’s likely that for many types of searches, the classic Search Box that returns a list of relevant web pages will still be the best experience - we do think it becomes less prominent, and possibly even specific only to certain types of queries.
Bottom line: As the search industry in general shifts to generative answers, we simply don’t see a win here for content publishers in the medium to long term. Barring unforeseen advancements, it appears to be all bad news.
PART 3:
While Google might be willing to inflict some pain on its content ecosystem & Adsense partners during the SGE transition period, I think it’s going to be a very different story for its advertisers.
As mentioned above, given that 60% of Google’s revenue comes directly from search ads, it’s hard to imagine they’ll play fast & loose with their golden goose.
But what does that actually look like if they shift most queries to SGE?
Will ads be displayed top-fold, with generative content underneath? Will they be more prominently featured within generated responses?
We have some idea what this looks like for Ecommerce ads, as shown below, but that’s just one category of advertiser, and that’s usually a lot further down the funnel.
But what does this look like for things like software companies, financial services, and all the other types of industries that don’t have SKUs or otherwise a “buy now” business model?
Obviously, we have lots of questions. And until we know more, we can’t do much more than speculate.
However, if Google does indeed take the plunge and rolls the majority of their search ad inventory into the SGE format, here are some reasonable assumptions:
Last, zooming out, the main difference in outcomes for advertisers vs publishers is that we know - with certainty - that Google has far more incentive to make this work. Again, 60% of their revenues depend on it.
So they’re going to pull out all the stops.
Bottom line: SGE will be disruptive, but likely not catastrophic, for most search advertisers. (Google has about 168 Billion reasons to not screw it up).
By definition, the dot-com bubble of 2000 was the first of many internet hype-cycles, where startups, investors and tech users in general profoundly overestimated exactly how the internet would disrupt other industries - as well as how long it would take to happen.
But they weren’t wrong to think that it would, in fact, change everything. In fact it’s safe to say that they profoundly underestimated that part.
GenAI is just the latest innovation on the block, and the current hype cycle is all but predictable.
Most of today’s pioneers in the space (possibly including ChatGPT) will ultimately fail, or be absorbed by the eventual winners. But there’s no question that AI will utterly change how we use the internet, and possibly even how most of us earn a living.
So it would be unwise to disregard today’s proverbial “Netflixes” as mere fads, or unserious gadgets, as they emerge. (Just ask Blackberry and Blockbuster how that worked out for them.)
It appears that Google, at least, has properly registered GenAI as a credible threat. So it’s doubtful that negligence will be their downfall.
But - taking the Netflix example - what’s unclear is whether Google’s traditional Search Box is the proverbial DVD - destined for extinction?
Or does it go the way of Cable TV - where its best days are well behind it, but it still has a place in the overall ecosystem?
Our hunch is it’s the latter - but anything can happen.
Regardless, internet entrepreneurs should watch this very closely. Google is clearly scrambling to stick the landing on a trillion-dollar maneuver... and whatever the result, it’s going to impact the nature of how the internet works, for all of us.
Our advice?
If you rely on search traffic in any capacity - this is a good time to think about channel diversification, and prioritizing your owned audiences.
Search won’t “go away”, but there’s also a good chance it will never be the same.
%20162x162.jpg)
%201%20(1).svg){kind=logo}
.svg)
.png)
%20162x162.jpg)
.png)
