Google's generative AI search is another nail in the content coffin
Now with the roll-out of Google’s generative AI search widget, those declines will go from “this is scary” to “falling-off-a-cliff-to-imminent-death” as Google stops sending free traffic to cybernetically-enhanced search-optimised pages.
Google’s $31bn decision to blow up the open web by using its AI tech to generate answers at the top of search pages has infuriated content publishers and site owners.
Why would Google do it, and why now?
AI not the reason
While the rapid prominence of AI is certainly the accelerator, it is not the reason.
Google has been taking incremental steps over the past year to increase the prevalence of zero-click searches.
Consumers appear to prefer it, but it doesn’t explain why the company has decided to release their gen AI widget now and risk breaking the traditional digital advertising ecosystem.
I would bet this is an opportunistic choice as both competitive and regulatory pressures rise – and the prevalence of AI assistant tools gives them the cover to remake their structure.
No need to search through pesky blue links
The gen AI widgets effectively remove the 10 blue links that made Google’s initial fortune with an auto-generated text summary of the answer you are looking for.
Searching for a recipe for a tasty salsa to go with your burrito? Here you go, no need to search through pesky blue links. Want to know the depth of the Mariana Trench?
The answer is fully formed in your results. Importantly, at no stage will you need to leave the comforting embrace of Google and open another page.
The Traditional google proposition
A small detour as an explainer. The Google advertising business has traditionally been based on a fairly consistent proposition.
Search offers an effective way for publishers to get free distribution. In exchange, Google gets to sell advertising around the search request, and then in classic network effects manner, they offer services to the site owners to get paid for that same traffic as part of the Google Partner network.
In their filed financials they refer to these pay-outs as traffic acquisition costs (TAC).
Inadvertently, this has directly incentivised the poor content creation loop that has created the Made for Advertising (MFA) industry.
This industry sucks $14bn annually from advertiser’s budgets for low-quality impressions, revenue that would otherwise have gone to genuine local and global publishers to fund much-needed original reporting.
But the model worked well
However, barring the rise of MFA, this model worked well overall. In 2013 the retained revenue (TAC – Google Partner Revenue) from the network was 34% (±$4bn) of the company's total net earnings, and it was effectively considered a good faith partner with publishers in their quest for a sustainable business model.
Wind the clock forward to 2023 however, and that ratio is now significantly less, closer to 12% ($9bn).
Google/Alphabet now pays more in distribution fees ($29bn) to a few select partners (hello Apple) than it does to the many thousands of Google Partner Network members. (We know who Google would prefer to keep happy.)
Search revenue has exploded
In the intervening period revenue from search and YouTube has exploded, new services have been added and the delta in network revenue has been substantially lower than on their owned and operated properties.
Net earnings went from $12bn to $73bn, so not only has the growth in earnings from the partner network stagnated in comparison, but it has also become significantly less material.
Clearly, Google could make this segment more profitable by simply increasing their take rate (the ±30% which they keep for managing and distributing advertising on behalf of advertisers and publishers).
However doing so in the current regulatory environment would lead to very public questions of the bamboo under the fingernails variety that no right-minded executive welcomes.
Why now
This leads us to a possible answer to the “why now” question.
Google is facing increasing pressure from regulators to unwind itself from its deeply integrated structure.
There are, quite rightly I believe, very strong arguments that through fielding players both on the buy and sell side of the digital advertising market as well as providing the referee (Google Analytics, the most widely used site measurement tool in the world) and renting the field (Google Cloud) they are simply overly vertically integrated.
In this situation, the units and technology that provide and manage the services for publishers could conceivably be the easiest ones to be hived off to appease the regulators.
In doing so, Google loses some cash, but not a huge amount, and they get to concentrate solely on their owned properties without distraction, much in the same way Meta does.
The regulators have a win, Google very publicly concedes and everyone feels like they have done a good thing for consumers.
If this is their direction of travel, then there is no longer an incentive for Google to keep funnelling traffic or cash to the network.
An unlikely mass exodus
This does, of course, change the nature of the alignment between publishers, who are a simple
But in reality, it is very unlikely that there will be a mass exodus because, well, what are the alternatives for site owners?
Traffic arbitrage from Outbrain and Taboola costs money, social is hard to win consistently, and newsletters require actual useful content.
where does this leave us?
Where does this leave the ecosystem?
Possibly with fewer available cheap SEO-driven impressions for marketers to buy, which may drive up prices in the short term, as well as increased competition for placement on search pages, which will give Google a revenue bump far in excess of what they lose from the partner network.
For publishers, significantly less free traffic directly affects site monetisation, which will have pretty dire consequences for genuine news and information publishers that rely on bulk traffic to grow short-term revenue and longer-term discovery and adoption.
Darwinian evolution to survive
For those who have made a business out of commissioning and optimising MFA content, I suspect there will be an enforced Darwinian evolution to survive.
They will attempt to game the new system, as they have done in the past, or simply use AI tools to create even more crap content across ever wider segments.
Without active management by marketers, this segment will continue to suck the oxygen from legitimate creators.
Another nail in an already nail-rich coffin
By exploding the methods of distribution and discovery, Google hasn’t made the internet a cleaner, better place, and the quality of the results from its AI tools is questionable.
However, it does become more profitable with fewer regulatory headaches, pleasing shareholders. And ultimately, a more useful destination for consumers.
For creators though - this is another nail in an already nail-rich coffin.
