LLMs, CHATBOTS
OpenAI’s relationship with agencies
OpenAI is trying to cut deals on its tools and services with agencies, but it’s not the cash cow the LLM-maker may have first imagined.
Adweek reported yesterday:
“In at least one recent negotiation, OpenAI sought a $1 million upfront commitmentin exchange for annual joint projects and expanded tool licenses, according to a holding company executive who requested anonymity to protect industry ties.”
“‘OpenAI is trying to find the right commercial model,’ the executive said. ‘[But asking for $1 million upfront], that’s not how you build relationships.’”
Adweek also provided detail on how agencies existing backend relationships with Microsoft – such as the Azure cloud – may be stymieing negotiations.
Read it. (July 8)
Thought bubbles:
- Though this is more of an enterprise story than an advertising and marketing story, this could foreshadow a complicated future for OpenAI’s in its relationship with agencies when the company makes its inevitable turn to advertising as a revenue source.
- The meta (but not Meta) here is that Microsoft and OpenAI are partners – and they are competitors, too. The agency landscape is just one field of “battle” for the AI titans.
- It’s ironic that agencies are being asked to foot a huge bill for AI while AI disrupts – some would say “destroys” – the business under their collective feet.
SELL-SIDE
AI Overviews and EU publishers
“Generative AI, not ad tech, is the new antitrust battleground for Google” (July 8) – Digiday (subscription)
-
“Independent publishers say Google’s AI Overviews have left them in a no-win bind — and they’re taking the fight to EU regulators.”
More: Google Hit With EU Antitrust ComplaintOver AI Overviews (July 7) – CNET
BRANDS
the AI competitive edge
Business of Fashion (BoF), a beauty trade, published a growth marketing “playbook” for brands about proper use of AI, data and moreyesterday.
Focusing on product creation as an element of marketing, and incorporating AI techniques, BoF saw faster ways for brands to respond to a feedback loop with the customer:
“… Brands need large volumes of regularly refreshed behavioural data. Tracking how people interact with content or describe products can help shape more accurate personas.”
”AI tools can accelerate this process, scanning reviews and comments to identify common desires or pain points. Jones Road used this tactic when launching a lightweight foundation, scraping online feedback to uncover priorities like sweat resistance or a natural finish. These insights helped the brand refine both product design and marketing, making the campaign more attuned to real consumer needs.” – Business of Fashion, July 8, 2025
Nevertheless, in an infographic for readers, Business of Fashion urged caution with AI even though it offers a competitive edge.
Read “From AI to Retail, Beauty’s New Marketing Playbook.” (July 8 – email required)
BRANDS
Podcast: The new chatbot challenge
An eMarketer podcast titled, “AI is Turning Bank Marketing Upside Down: What It Will Disrupt First, Second, and Then Third,” eMarketer analysts discussed if AI has already changed FinServ’s approach to marketing as well as what the future holds.
Podcast host Rob Rubin pointed out that eMarketer’s forecast for US Financial Services digital ad spend for 2025 is $42.6 billion.
eMarketer analyst Lauren Ashcraft responded that she saw a new marketing challenge brought on by chatbots:
“I do think there’s some possibilities with allowing better understanding of customers through [AI], which can feed back into data-driven marketing.”
”But what’s really about to change bank marketing,I think, is something that is a challenge. It’s being able to have your content show up in generative search engines because it’s a completely different game than having a great SEO score or showing up in regular search engines.”
“You have to tailor your content to answer questions that people are using generative search engines for. So that’s new. And maybe two months from now, there’s going to be completely new things that bank marketers have to keep thinking about. So it’s just constantly changing. That’s how AI is impacting bank marketing right now.”
Hear more on the Apple Podcasts app (July 8).
BRANDS
Unilever plunging into AI and creative
After Monday’s news about Unilever’s AI-driven design unit, the global brand marketer was singing the praises of AI againyesterday and its ability to garner millions of impressions with consumers via social media for its beauty care products.
The Wall Street Journal reported:
“‘AI was critical to getting those impressions,’ said Ryu Yokoi, chief media and marketing capability officer, Unilever North America. The company took over 100 discrete pieces of influencer content such as stills and short clips and used generative AI to remix them into different sizes, formats and lengths to tailor it to different audiences on different social-media platforms, he said.”
”’Unilever is seeing proven value in working with human influencers, but one day AI-generated influencers might play a role,’ according to Yokoi.”
”’AI is going to impact so many different aspects of this,” he said. “I think it depends on the brand and the use.”
Read more in the WSJ. (July 8)
TECH
The effects of AI mode
In a story about changes in Google ads data after the implementation of Google’s AI Mode in Search, Search Engine Land reporter Jason Tabeling observed yesterday, “Google hasn’t done a full cannonball into the large language model (LLM) pool, but they’re definitely in the water.”
This article from a search marketer’s viewpoint is loaded with statistics as the changes – and AI’s effects – are subtle sometimes and other times not. For example, Tabeling wrote:
“I looked at throughput metrics like click-through rate and conversion rates. These data points paint a very interesting picture. Across the board, CTRs have decreased fairly sharply. Every single category is down since January.”
”Shorter keywords have a CTR drop of 50%.”
”For longer keywords, the decrease is less dramatic, but only slightly.”
“Keywords with more than eight words are down by 26%.”
Image from Search Engine Land
Read the tea leaves of Google Search post AI Mode introduction. (July 8)
TECH
Podcast: Agents for programmatic rails
On the AdTechGod podcast, AdTechGod speaks with Taylor Simons, a former MediaMath executive and founder of programmatic consulting services company, TCHT.
Simons provided insights on a variety of topics (why there are significantly more SSPs than DSPs) including AI where Mr. Simons sees opportunity with AI agents:
“… when I look at what’s happening with AI agents, I actually very strongly believe that we can make this a new channel and/or a new format for the programmatic open Internet.”
”We haven’t had a new channel – or a new format – since CTV. We saw what growth that brought.”
“I was speaking with a few different C-level executives at [DMS – Digital Media Summit] a couple of weeks ago on this very topic.”
”I want to be at the forefront – and I want my clients to be at the forefront – of enabling, let’s say, text ads and new search ads within these AI agents. That’s changing dramatically and I think it’s a very real thing and we can make it happen on the existing programmatic rails.”
Hear more on the Apple Podcasts app. (July 8)
TECH
Hiring for neuro-contextual AI
A Seedtag company blog post said yesterday that Brian Gleason, who most recently helped guide Criteo’s commerce media business, will become CEO of the AI ad tech startup while co-founders Jorge Poyatos and Albert Nieto “will transition to new roles as Chief Innovation Officer and Strategic Advisor, respectively.” Both are ex-Googlers. Read more.
Seedtag positions itself as an omni-channel advertising solution using “neuro-contextual AI.” Mr. Gleason, who is already hitting the ground running, told Adweek yesterday that his company’s “tech uses AI to assess the emotion invoked by content on a webpage or CTV streaming platform in real-time and then uses that information to place ads based on a brand’s product category and campaign objectives….” Read that one.
TECH
Adobe on Amazon AI
- Adobe Says Retailers’ AI-Driven TrafficWill Leap 3,200% During Prime Day (July 8 ) – PYMNTS
- Prime Day event expected to drive $23.8 billion in online spend across U.S. retailers (July 7) – Adobe
TECH
Podcast: Advertising mega-trends
Promoting his upcoming book, “Yield: How Google Bought, Built, and Bullied Its Way to Advertising Dominance,” Marketecture CEO Ari Paparo appeared on Aqfer CEO Dan Jaye’s new podcast and discussed a range of AI-related topics.
Mr. Paparo’s “mega-trends” to watch out for in the year ahead are:
- Meta/Google.. what are they gonna do with AI/ads? – specifically their “one button” campaign manager…
- AI adoption among agencies– how big of a change is this for their business models?
- Finally, can AI be “modular” or must it be “native”when it comes to ad tech. Do you need to build from the ground up to be effective in advertising in the age of AI?
Listen to the MADTech Momentum podcast titled “Ari Paparo: AdTech, AI, and the Battle for Attention” on Apple Podcasts app. (July 8 )
AGENCIES
Creative AI machina
“This Company Says It’s Found A Way To Balance Automation With Human Control” (July 7) – AdExchanger
- Making Science, a 1200-person, digital marketing consulting firm has an agentic AI platform called “ad machina” and is based in Madrid.
- The firm “recently launched a tool called Creative Hub that helps advertisers optimize their marketing campaigns by bridging the gap between creative and execution while also providing brands with full control over their creative versioning.”
- Making Science Launches Creative Hub, a Performance AI Tool that Boosts Ad Sales by Up to 20% (June 19) – press release
PROMPT
Prompt: How can web publishers overcome the loss of web traffic due to users use of LLM chatbots?
Response from Anthropic’sClaude.ai:
Web publishers are facing a significant challenge as LLM chatbots increasingly provide direct answers to user queries, reducing the need for users to visit original websites. Here are the key strategies publishers can employ to adapt:
Answer Engine Optimization (AEO): Publishers need to evolve beyond traditional SEO to optimize for AI systems. This means structuring content to directly answer common questions in clear, scannable formats that AI chatbots can easily parse and cite. Content should be authoritative and well-sourced, making it more likely to be referenced by AI systems even if users don’t click through.
Create Irreplaceable Content: Focus on content that AI chatbots cannot replicate or summarize effectively:
- Original reporting and breaking news
- Exclusive interviews and expert analysis
- Local coverage and community-specific content
- Interactive experiences and multimedia content
- Real-time updates and live coverage
- Personal stories and human-interest pieces…
HR
You’re hired
- Ken Rona is joining JWP Connatix (was JW Player) as its first Chief AI Officer. Mr. Rona said on LinkedIn yesterday: “New job announced! I am working for John Nardone (who was my CEO at [x+1]) at JWPC. They were consulting client and we decided to work together, full time. Going to go into the office for the first time in just a bit. But thanks to all the JWPC folks for making me feel welcome over the last 3 months. Short version is that I am responsible for evangelizing AI, both internally and externally, rapid prototyping using AI, and the ML team. The plan is to be an AI-first company and it was an amazing opportunity to learn and work for someone who a top notch CEO and help build some really innovative products.” – Kenneth Rona on LinkedIn
- Andrew Pierce announced he joined as Instacart Ads Senior Product Manager(was Meta) as of February (July 8 ) – LinkedIn
- Former TripleLift CEO and FreeWheel GMDave Clark joins Cast Iron – LinkedIn
MORE
- Opinion:How marketers can build cultural fluency in the age of AI (July 8 ) – Ad Age (subscription)
- Report: Mistral Aims to RaiseAs Much As $1 Billion in Financing for AI Cloud Service (July 8 ) – PYMNTS
- AnyMind brings agentic AIto influencer workflows with latest AnyTag update (July 8 ) – press release
- AI ‘likely’ to produce bestseller by 2030, Nielsen’s Philip Stone tells Marketing & Publicity Conference (June 30) – The Bookseller