Clocking Out<\/p>\n
Looking for a long read on the behind-the-scenes maneuvering that has overhauled Meta in recent years?<\/p>\n
Reuters caps off its years-long series on Meta\u2019s AI investments and employee layoffs with a deep dive<\/a> into Project OT (for \u201cOrganizational Transformation\u201d). That\u2019s the internal shorthand for Meta\u2019s plan to greatly reduce headcount while offloading services to agentic tech. (The initiative follows Meta\u2019s last major overhaul-gone-bust, when it went all in on the metaverse.)<\/p>\n The story will ring painful bells for Meta advertisers, who felt the platform\u2019s customer service slip as account services were handed off to AI.<\/p>\n Individual advertisers have lost millions and even tens of millions of dollars when Meta\u2019s ad platform glitches out or misappropriates campaign budgets<\/a> \u2013 and countless billions of dollars have evaporated overall in the past four years. Advertisers aren\u2019t even informed<\/a> when glitches devour their budgets. There are no people left to inform them. Agencies that spend hundreds of millions of dollars on Meta can only turn to a useless chatbot that asserts all is well.\u00a0<\/p>\n You might also recall, in 2022, some Meta employees were quietly fired for accepting cash kickbacks<\/a> from advertisers simply to have normal customer service tickets seen to. That\u2019s probably not what Meta had in mind when it comes to AI creating new revenue streams.<\/p>\n Ask And You Shall Receive<\/p>\n Every tech company these days seems to be developing a centralized AI interface. Next up: The Trade Desk.<\/p>\n TTD\u2019s new agentic system, called \u201cAsk Koa,\u201d serves as a \u201csingle conversational entry point to a range of AI agents,\u201d per Adweek<\/a>. The agentic chatbots handle different parts of the programmatic workflow, like audience building, optimization and generating campaign insights.<\/p>\n According to a leaked product road map obtained by Adweek, the \u201cAsk Koa\u201d interface will move from alpha to closed beta this week, with plans to expand further in the coming months.<\/p>\n Another key feature of the road map is a frequency capping tool that automatically adjusts how often a person sees an ad based on the brand\u2019s performance data and campaign pacing.<\/p>\n After a less-than-enthusiastic reception to the shift to Kokai, TTD needs to win back customer trust. Not to mention winning back customers themselves after many advertisers purportedly shifted some of their programmatic budgets to Amazon DSP<\/a>.<\/p>\n \u201cA large majority of the input to our product runway process comes from customer feedback,\u201d says Eric Bosco, TTD\u2019s VP of product management.\u00a0<\/p>\n Reduce, Reuse, Restructure<\/p>\n USA Today Co. is the latest in a long line of publishers \u2013 including Time and The Economist \u2013 restructuring its content for AI bots.<\/p>\n The new formatting will make it easier for AI systems to process and cite USA Today\u2019s content, Kara Chiles, SVP of product management, tells Digiday<\/a>, and is part of the publisher\u2019s larger GEO strategy. USA Today has a strict whitelist of approved crawlers, and all others are blocked by default.<\/p>\n These changes won\u2019t have much of an impact on human site viewers; in fact, said Chiles, most of the modifications won\u2019t even be visible to humans. Rather, she said, \u201cit\u2019s a lot of the infrastructure, the metadata, how we\u2019re actually structuring our formats and templates to be as discoverable for AEO and GEO.\u201d<\/p>\n But the exact formats haven\u2019t been determined yet. USA Today is still in the process of testing which content performs best among crawlers.<\/p>\n Ultimately, the pub\u2019s goal in revamping its content structure is to strike more licensing deals with trusted AI companies and provide more value within its current deals.<\/p>\n Chalk it up as more proof publishers are increasingly tailoring their content to bots, not just humans.<\/p>\n But Wait! There\u2019s More!<\/p>\n Meta agrees to pay up to $18 billion over 10 years to 48 states to settle a federal lawsuit alleging its social media platforms harm children and teens. With the settlement, which is the single largest-ever Big Tech legal payout in history, Meta admits no wrongdoing and avoids a potential $200 billion penalty. [WSJ<\/a>]<\/p>\n Attaching a celebrity\u2019s name to a brand isn\u2019t a replacement for good marketing. [Modern Retail<\/a>]<\/p>\n Reddit is no longer a leading source in ChatGPT citations \u2013 but that doesn\u2019t mean marketers should stop caring about it. [Business Insider<\/a>]<\/p>\n AI slop is so prevalent online that even cute animal pics can\u2019t be trusted. [Wired<\/a>]\u00a0<\/p>\n OpenAI is testing a new feature that allows advertisers to call out the types of content they don\u2019t want to advertise against. [Adweek<\/a>]<\/p>\n X sent out a cease-and-desist letter demanding that companies, including Nitter and XCancel \u2013 which scrape content from X so people can see it without actually accessing X\u2019s own site \u2013 be shut down. [Forbes<\/a>]<\/p>\n