AI advertising platform StackAdapt introduced Ivy Studio, an AI-first workspace that replaces traditional dashboards with natural language for campaign planning, analysis, optimization and execution.
A press release makes the case:
“Rather than asking marketers to learn software workflows, Ivy Studio enables marketers to define the outcomes they want to achieve, while AI agents grounded in StackAdapt’s advertising intelligence help uncover insights, identify opportunities, recommend next steps, and move work forward.
Built directly within the StackAdapt platform, Ivy Studio serves as a unified AI-first advertising hub for planning, forecasting, analysis, optimization, and execution. Leveraging AI agents, marketers can move from insight to action without switching between tools or workflows.”
StackAdapt also announced an MCP server that lets AI assistants such as Claude access campaign data and reporting programmatically outside the platform.
CEO and co-founder Vitaly Pecherskiy of StackAdapt said on LinkedIn yesterday, “You know you have something special on your hands when you see usage stats from beta testers going up and to the right.”
Read: StackAdapt Introduces Ivy Studio, a New Hub for AI-First Advertising – press release
From tipsheet: StackAdapt’s emphasis on advertising intelligence, live campaign data and MCP connectivity suggests the next generation of ad platforms could be judged less by their user interfaces than by the quality of context they make available to AI.
LLMs & CHATBOTS
Building the org
- “Thrilled to be joining OpenAI, leading Strategy & Operations for Business Marketing!” – Former Notion marketer (July 28) – Jamie Lynn White on LinkedIn
- “I’m so excited to share that I have joined OpenAI as a founding member of the GTM Ads team, where I will be standing up our Telco & Professional Services verticals.” (July 28) – Former Meta Industry Manager Katie Baker on LinkedIn
More: “OpenAI is already building the org chart of a mature ad business” (July 28) – Digiday
From tipsheet: OpenAI’s advertising strategy is becoming visible through its org chart before it is visible in its products.
Related: OpenAI’s Ads operating model emerges (July 21) – tipsheet
FINANCIALS
Earnings reported yesterday
Public companies are reporting their latest quarterly financials and providing a look into how AI may be creating opportunity for the advertising and marketing ecosystem.
Omnicom
- Markets: Omnicom stock was down 3% in after-hours trading yesterday.
- Stat: “2026 Second Quarter – Core Operations: Revenue of $6.0 billion, 6.1% organic growth; Non-GAAP Adjusted EBITA of $1.1 billion, 17.8% margin”
- Company: Omnicom Reports Second Quarter 2026 Results (July 28) – Omnicom (PDF)
From tipsheet: This earnings season highlights different visions for competitive advantage in the AI era among agency holding companies. Publicis argues scale is becoming more valuable because of AI. Havas argues AI reduces the importance of scale by rewarding agility. Omnicom is betting the winning formula is greater scale combined with proprietary data, identity and AI.
Note: WPP reports its interim results on August 6.
AGENTS
Google’s Buyer Direct gains momentum
Coming on the heels of Gamera CEO Gareth Glaser’s op-ed about Google’s Buyer Direct product on AdExchanger and the discussion it ignited (covered by tipsheet yesterday), a Google executive highlighted early results from the offering on Monday.
In a LinkedIn post, Paul Gubbins, CTV Lead, EMEA for Google Exchange Platforms said:
“By pairing the best features of traditional insertion orders with the advanced controls of Programmatic Guaranteed, it is now possible to buy directly from Google Ad Manager.The results?
- -50% fees
- 5 minutes setup time
- Several clicks to launch the campaign (compared to more than 20 in classic PG)”
Referencing the challenges of supply path optimization for agencies, he acknowledged client partners involved in the Buyer Direct test (see below) including WPP Media, Prisma Media and Nestlé.
Read more. (July 27)
Case study: “Buyer solutions success stories” – Google Authorized Buyers Help
From tipsheet: Google’s messaging is shifting from explaining Buyer Direct to selling it. The company is emphasizing lower fees, simpler execution and supply path optimization while pointing to recognizable agency and advertiser partners. That suggests Buyer Direct is moving from product introduction toward commercial rollout.
PROTOCOLS
MCP’s biggest rewrite yet
The Model Context Protocol (MCP) has published its biggest update since its November 2024 launch.
On LinkedIn yesterday, Aqfer’s Daniel Jaye outlined the significance of the new features:
“MCP just shipped the biggest rewrite of its protocol since launch. Ad tech should be paying attention. Quick context if you’re not deep in the Model Context Protocol (MCP): it’s the emerging standard that lets AI agents call tools and data sources in a consistent, secure way — think of it as the plumbing that lets an LLM safely query a database, trigger a workflow, or pull live data, instead of everyone building bespoke integrations. It’s becoming the default way enterprises connect AI agents to real systems.”
Jaye argues the changes validate Aqfer‘s architecture and advance the company’s vision for developing MCP extensions for advertising and marketing technology.
Read more. (July 28)
Read: “The 2026-07-28 MCP Specification Release Candidate” (July 28) – Model Context Protocol blog
From tipsheet: MCP is evolving from a tool-calling protocol into enterprise infrastructure. As AI agents move into production, standardized interfaces for connecting models to enterprise systems could become as important as the models themselves.
LLMs & CHATBOTS
Developments
- Opinion: “The AI Future Is for Everyone” (July 28) – Mark Zuckerberg, CEO, Meta in The Wall Street Journal (subscription)
- “Your Claude shared chats and Artifacts may have ended up on Google” (July 27) – TechCrunch
- “Today, we’re opening applications for the first cohort of OpenAI Campus Leads – a global program.” (July 28) – Leah Belsky, VP Education, OpenAI on LinkedIn
MEASUREMENT
AI slop outperforms quality metrics
A new report from the Trustworthy Accountability Group (TAG), the Association of National Advertisers (ANA), and data management platform Fiducia finds AI slop is getting better at mimicking legitimate inventory
The findings cover only open web programmatic. Nevertheless, Adweek’s Trishla Ostwal writes:
“In some cases, advertisers are paying more to appear next to AI slop than they would to appear on legitimate publishers. The problem spans open web and social media platforms, which, the report claims, is the fastest-growing environment of AI slop. AI slop accounts for 1.3% to 2.4% [of open web programmatic ad spend]…”
Read: AI Slop Is Fooling Advertisers’ Verification Tools (July 28) – Adweek (subscription)
According to the study, AI slop recorded lower invalid traffic, higher viewability and higher TrueCPMs than clean inventory, allowing low-value content to outperform legitimate publishers on many of the signals advertisers use to evaluate quality.
More:
- Press release: “TAG/ANA/ Fiducia Analysis Quantifies Level of “AI Slop” in Digital Advertising Supply Chain for First Time” (July 28) – read more
- Get the study: “TAG, ANA, Fiducia Study Finds AI Slop Accounts for Up to 2.4% of Open Web Programmatic Ad Spend” (July 28) – TAG Today
From tipsheet: AI is optimizing for advertiser quality metrics rather than reader value.
Remember when: One old trick among disreputable publishers was to build pages that existed primarily to generate ad impressions, or clicks, rather than satisfy readers. AI slop is the next evolution, using AI to optimize for the very metrics advertisers use to evaluate quality.
AGENTS
Hightouch scaling creative with agents
HelloFresh marketer Conor Feeney tells AdExchanger that his meal-kit company needs creative that not only performs better but can be produced, tested and prioritized much faster. To do that, the company turned to AI and Hightouch‘s agentic marketing platform.
AdExchanger’s Joanna Gerber reports:
“HelloFresh adopted Ad Studio to meet this need for creative at scale. Ad Studio (which is one tool in Hightouch’s larger agentic marketing suite launched earlier this year) handles the creative process from asset generation to deployment across platforms… One of the biggest challenges in developing AI creative is ensuring that a brand’s voice doesn’t get lost or misrepresented, and that all of its content regulations are met. Hightouch addresses this by using a brand’s own content (including product catalogues, legal and compliance guidelines and previous ads) for guidance.”
Read: How HelloFresh Brought AI Creative To Times Square (July 28) – AdExchanger
More: Hightouch Launches Ad Studio to Create On-Brand Ads at Scale with AI Agents (April 16) – press release
From tipsheet: Hightouch’s “Brand Context Layer” grounds foundation models in a company’s own creative assets, campaign history, brand guidelines and performance data. That lets marketers generate more creative, iterate faster and continuously optimize while staying on brand.
Related: “A Company Is Producing 1,200 AI Ads a Day. It Reveals a Hard Truth for Every Brand” (July 27) – Inc.
COMMERCE MEDIA
Retailers prepare for AI-referred shoppers
Retailers are beginning to think beyond visibility in AI-driven product discovery and toward what happens after the recommendation.
Criteo President of Retail Media Sherry Smith said during a recent panel discussion with The Drum at Cannes that while generative AI can sometimes get things wrong, retailers hold a more durable advantage: first-party customer relationships. Rather than competing with AI, she suggested retailers should “power up” AI-referred traffic using purchase history, loyalty data and customer preferences to create more personalized shopping experiences.
Albertsons SVP of Digital Customer Experience Jill Pavlovich described agentic shopping as “a blend of human and machine,” arguing that retailers should let customer behavior shape how AI is integrated into commerce. Google and Mars United executives added that success will still depend on strong product content, digital shelf execution and closer collaboration between brands and retailers.
See the video and read: “‘How much LLMs can lie to you’: why retailers need to own the AI shopper” (July 28) – The Drum
From tipsheet: AI and conversational interfaces are shifting the point of product discovery, making retailers’ proprietary data more valuable than ever. Purchase history, loyalty data and merchandising intelligence enable retailers to personalize every interaction. AI may generate the recommendation, but retailers own the customer relationship and determine whether it becomes a repeat purchase.
TECH
Podcasts and ads
- Why Is the Podcast Ad Market So Messy? (July 27) – Ari Paparo on Marketecture
- RedCircle Expands AI Tool To Screen Programmatic Podcast Ads For Brand Safety. (July 28) – Podcast News Daily
- Study: “AI Audio Could Take Over Radio, Podcasts And Audiobooks” (July 22) – Frank Racioppi on Forbes
MORE
- Inside LiveRamp’s OpenAI Partnership: A Q&A With Luke Fenney, SVP of Publishers and Platforms (July 20) – Futureweek
- Higgsfield AI Launches $85,000 ‘Adathon’ Contest in Partnership with Adweek (July 27) – Adweek
- How Adobe’s Marketing Team Treats AI as Infrastructure, Not a Campaign (July 27) – Adobe on LinkedIn
- Opinion: “Ad Tech’s New Standard Is a Vector With No Ground Under It” (July 28) – Nate Woodman, Proof in data on Substack
- How AI Is Rewriting The Economics Of Streaming Advertising (July 28) – Forbes