• by Joe Mandese , 7 hours ago

Ever since the Big Bang of generative AI three years ago, ad industry doomsayers have been predicting the deprecation -- if not the demise -- of the search advertising marketplace. And while generative AI has been nipping at paid search's buds, it has not proven so far to be the dystopian search advertising scenario that some have been predicting.

While it certainly is beginning to impact search's share of ad spending, a more likely scenario is that the nature of search advertising is evolving that is blending conventional search with generative AI-based queries and recommendations. It will be interesting to watch how that plays out over time.

In the meantime, generative AI appears to have a new victim -- the open web in general, and open web publishing in particular.

Experts have been speculating on and citing anecdotal evidence that so-called AI "overviews," the AI-powered summaries popping up in Google's and other search engine results, have begun eating into clickthroughs -- and ad share -- of open web publishers that previously had been deriving much of their traffic from Google and other search engine queries.

The logic: If someone can get a concise, well-organized gist of information in an overview that they normally would have had to click through to a publisher's site to read for themselves, why would they bother.

This week, the ad industry economists at Madison and Wall published a series of second-quarter ad spending analyses that I believe may be the first material ad marketplace proof that open web publishers are indeed the canaries in an AI-powered coal mine.

After publishing an update showing a nearly 16% increase in digital ad spending helped drive double-digit growth in total U.S. ad spending, the Substack newsletter followed up with a drill-down focusing explicitly on the open web, as well as its sub-components, revealing surprisingly tepid growth in the quarter.

While not yet declining like most of the non-digital ad-supported media have, Madison and Wall's analysis finds total open web ad spending expanded just 3.4% -- or about a third of total U.S. ad spending growth during the quarter.

Even more alarming, the analysis shows that open web publishers -- including many of the sources of news and information that Americans have historically counted on for staying informed -- plummeted to just 0.5% growth.

"Notwithstanding the vital importance of independent journalism and the web’s critical role in supplying it, we continue to maintain a pessimistic view on the future of the open web as an advertising business," this morning's edition of Madison and Wall notes in a dispatch that make no mention of the impact of AI, much less AI overviews, in driving traffic and ad revenue to publishers.

Needless to say, the development is yet another factor in a series of economic as well as political factors threatening the vitality of publishers.

Publishers have been fighting back -- mainly by calling foul, finding ways of blocking AI overviews of their content, including legal action -- which may prove to be a Pyrrhic victory in the long run for the simple reason that if they no longer show up in query results (via conventional search, AI search or even AI overviews) their content will grow increasingly inaccessible and relevant to many web/AI users.

"Eighty-three percent of newspaper and magazine websites globally now block AI bots... using server-side controls and anti-bot technology," according to a press pitch I received this morning citing new research from cybersecurity firm ImmuniWeb.

I haven't reviewed the research in its entirety yet, but it seems about right based on what I gleaned from glancing at it.

I'm not sure what a better, long-term solution is for publishers -- especially the vital journalism kind -- but predictions of alternative business models so far have not really emerged to replace their historic revenue streams. You know, the progression that began ever since Jeff Zucker's infamous quote that the web was converting "analog dollars into digital dimes."

I hope Zucker doesn't mind if I update it with a new quote that AI overviews are converting "digital times into AI pennies, if even that."

While leading AI developers have been cutting some licensing deals with publishers to legitimately access their content to train their large language models and provide indirect access to users, it remains unclear how those models will evolve over time -- and more importantly, whether it will ultimately sustain the direct relevance publishers have with their end users. You know, what we used to call "subscribers," "readers," and maybe even "informed citizenry."

That said, it is possible that the next generation of AI could also facilitate one of the long-vaunted business models of so-called "micropayments" that publishers and long-term industry forecasters have been touting as a means of offsetting losses in advertising and conventional subscription models.

In fact, WPP Media devotes an entire section to it in its just-released "Advertising In 2030" update on a benchmark report first published five years ago.

Sadly, WPP Media's take is that it's actually moving in the wrong direction of micropayment probability.

You can read it yourself in section 12 of the new report, but the bottom line, according a panel of 60 industry experts surveyed by WPP Media, is:

"Experts are more pessimistic about this now than in 2020, with 67.7% viewing it as unlikely. Friction, both practical and psychological, has been and will remain the main blocker," the report finds, adding: "For publishers, advertising still pays the bill for many. And on the premium end, old-fashioned subscriptions are still the way to go."

Good luck with that.