Here’s today’s AdExchanger.com news round-up… Want it by email? Sign up here.
Pour One Out For Pub Tech
During the heady days of gonzo social traffic and zero-dollar interest rates, publishers were launching their own software businesses.
But that cottage industry has practically disappeared without so much as a whimper.
Vox Media, one of the keenest publishers to launch tech, will stop using its homegrown CMS, called Chorus, Axios reports.
Gawker tried and failed, too. Even The Washington Post, with better resources and Jeff Bezos to spur tech investments, stopped licensing its ad tech and publishing software, Zeus, last year.
There are counterexamples to the “media can’t do tech” narrative. But they’re mostly mar tech companies with plug-in media – and the publications aren’t major news destinations.
Minute Media is roughly split between sports sites and ad tech. System1 is a tech-and-media hybrid, as is The Arena Group, which collects former magazines with name recognition, like Sports Illustrated, TheStreet and Men’s Journal.
Never count out another publisher tech hype cycle, though.
“The next iteration of Zeus is part of our future story around ad targeting and brand suitability,” WaPo’s new chief advertising officer Johanna Mayer-Jones recently told AdExchanger about the defunct software business.
License To Crawl
It’s no secret: Big Tech’s large language models (LLMs) use publisher content to train their algorithms.
Publishers can block crawlers from accessing their pages by editing their robots.txt files, but this approach is flawed, writes SEO expert Fili Wiese in an op-ed for Search Engine Land.
For one thing, some publishers don’t even have permission from their web servers to edit robots.txt files.
Companies that own LLMs – namely, Google and Microsoft – also operate search engines. So blocking LLM access would also prevent the publisher from indexing for search traffic, which they do want.
Instead, Wiese proposes that LLM operators and publishers use Creative Commons licenses, which are already used online. The meta tags in an HTML file could indicate whether usage rights have been granted to LLMs and/or search engines.
This way, publishers can block LLM crawlers, if they choose, without cutting off search crawlers.
The burden would then shift to the tech companies, which would have to limit themselves and be more cautious with content for training machine-learning systems.
Cloudy With A Chance Of Pay Me
The word is that Microsoft’s generative AI products for its corporate software suite will cost a pretty penny, Bloomberg reports.
Why does that matter to ad tech and media?
Because it’s a harbinger of the immense costs in cloud computing that will soon descend upon the entire industry.
The big three cloud players – Amazon, Microsoft and Google – all back their own generative-AI startups. Microsoft uses OpenAI, of ChatGPT fame, while Google invested in Runway at a $1.5 billion valuation. AWS has a $100 million fund for generative AI businesses.
That’s because these machine-learning software services are, unsurprisingly, huge data consumers. A generative AI startup practically has to cut a deal with one of the big cloud providers because they need the computing power.
In ecommerce and DTC marketing, Google, Meta and Amazon are often framed as landlords: Brands pay them rent or are in the gutter. Soon, cloud providers like AWS, Azure, GCP and Snowflake may very well take over that role in ad tech and digital media.
But Wait, There’s More!
In praise of phone numbers. [The Atlantic]
IPG Mediabrands creates a unified retail media solution. [release]
NBCUniversal’s Peacock announces its first price hike since launch. [TechCrunch]
A blessing and a boogeyman: Advertisers warily embrace AI. [NYT]
The holding company Stagwell acquires Tinsel Experiential Design. [release]
Norway is the latest country to pressure Meta’s ad business. [Digiday]
You’re Hired!
Wavemaker names Premjeet Sodhi as global head of measurement and analytics. [MediaPost]
