The Digiday Podcast: "Why AI is Agencies’ Frenemy"

Date: August 26, 2025
Hosts: Kamiko McCoy (A), Tim Peterson (B)
Guests: Seb Joseph (C), Michael Berge (D)

Episode Overview

This episode of The Digiday Podcast dives deep into the ways artificial intelligence (AI) is upending the advertising agency model—making AI both a formidable disruptor and a necessary ally for agencies. Through a lively, insightful discussion among Digiday editors and reporters, the episode explores:

• Recent AI industry drama between tech titans
• The potential bursting of the retail media bubble
• A new wave of compensation models for both publishers and agencies dealing with AI
• How agencies are grappling with the threat (and promise) of in-housing and automation through AI
• The future of entry-level jobs and the agency holding company model in the age of agentic AI

Key Discussion Points & Insights

1. AI Legal Drama and Platform Competition

[04:40–11:22]

• Elon Musk’s X.AI vs. Apple & OpenAI:
  Elon Musk’s X.AI has sued Apple and OpenAI, accusing them of conspiring to block competition in artificial intelligence. The core allegation claimed Apple’s partnership with OpenAI prevents the X app and Grok app from App Store prominence.
  • Quote (B, 05:50): “If you can’t beat them, sue them. But that really does seem to be this man’s philosophy here... this is a regular behavior from Musk where if my business is not winning, boom. Lawsuit.”
• Reality Check on App Store Rankings:
  Rankings are download-based, not partnership-based. ChatGPT’s dethronement by Hank Green’s Focus Friend app illustrates the openness of competition.
• Meta-debate on AI Platforms:
  Grok, X’s AI, isn’t a major player compared with established models like ChatGPT and Google’s Gemini, raising doubts about Musk’s claims.
• Apple’s AI Dabbling is Wide:
  Apple is now also in talks with Google about using Gemini to power Siri, weakening allegations of “exclusive” relationships.
  • Quote (A, 10:58): “I don’t know how strong of a case you have ... if you had a bean, a fake bean, dethroning ChatGPT and Apple in talks with Google.”

2. Retail Media’s Inflection Point

[11:22–15:40]

• Retail Media Booming Yet in Jeopardy:
  With major retailers like Macy’s joining Amazon’s ad network and smaller players scrambling for tech, the retail media boom may be at risk if sales soften.
• Disparity Between Retail Media Revenue & Product Sales:
  Retail media revenue is surging (e.g., Target: $217 million vs. $162 million YoY; Walmart Connect: +31%), while actual retail sales stay flat, raising questioned value for advertisers.
  • Quote (B, 13:54): “Product sales not doing so hot, retail media sales doing hot, but if the retail media sales are to drive product sales and the product sales aren’t happening… in nine to 12 months, are we going to see a drop in the retail media sales?”
• Advertisers Demanding Accountability:
  Buyers now have data to challenge ROI claims from retail media, which may force a reckoning in how impact is measured and compensation structured.

3. Shifting Publisher Compensation Models for AI

[15:40–20:19]

• The Shortcomings of Content Licensing Deals:
  Publishers have licensed content to AI companies for one-time fees, realizing these aren’t sustainable or growth-oriented.
  • Quote (B, 15:40): “The value isn’t really there for the publishers long term ... this is not the way forward.”
• The Rise of Usage-based Compensation:
  New frameworks (like those from IAB Tech Lab) and real-world deals (e.g., Perplexity’s ad revenue-sharing with Gannett) are moving toward usage-based or subscription-split models (e.g., Comet Plus shares 80% of $5/month with publishers).
  • Quote (A, 19:46): “At least you have more of a say ... it does bring up the conversation and what this compensation model could look like.”

4. Feature Segment: Agencies’ AI Compensation Crisis

[21:05–51:12]

A. AI In-housing (or not): Realities for Brands

• Varied Receptivity Across Brands:
  Some, like Adobe, build AI tools in-house to sell them onward, while others (e.g., US Bank) balk at the cost and complexity, preferring to outsource.

  • Quote (A, 22:01): “It comes down to data privacy and how much cash I’m willing to spend.”

• The Programmatic Precedent:
  This echoes the late 2010s, when brands dabbled with bringing programmatic buying in-house, only to retreat as complexity and costs mounted.

  • Quote (C, 24:13): “Advertisers go through that kind of Gartner hype cycle ... you get that kind of trigger point ... then the trough of disillusionment ... I’m bullish on the fact that I don’t think this will be a ... death knell for agencies ... more a transformation.”

B. Agencies Adapt, But Worry About Value

• Less Panic, More Adaptation:
  Agencies are less anxious about AI in-housing than they were with programmatic. The greater focus is how to adopt AI for their own benefit—and how to price it.

  • Quote (D, 27:50): “There’s less of that kind of jittery… response of ‘Oh my God, we’ve got to do something to stop this.’ Agencies are so busy trying to figure out how they can incorporate AI ... that’s kind of keeping them pretty occupied.”

• Agencies’ Added Value Lies in Tech Stacks & Data:
  Agencies can layer their own tools/data atop AI platforms, mitigating the risk of brands bypassing them directly.

• Distinctions Between Holding Companies and Independents:
  Holding companies, with scale and proprietary tech, are better positioned to withstand AI-driven disruption than smaller independents.

C. From Outputs to Outcomes: The Measurement Quandary

• Agentic AI’s Impact on Research and Planning:
  Agencies are using agentic AI to assemble virtual focus groups/personas, enabling faster (if yet unproven) consumer insights, both saving costs and changing workflows.

  • Quote (D, 34:06): “Agentic is being used ... to create kind of virtual focus groups ... glean these insights ... in a matter of minutes. Now, what I don’t know ... is how accurate these Personas are.”

• Value Measurement & Remuneration Evolve:
  AI enables outputs faster—thus threatening time-based compensation. Agencies are pressed to pivot to outcome-based or usage-based models, but measuring true impact remains murky.

  • Quote (A, 37:25): “If you’re charging based on how long something takes and you cut it down to 20 minutes, what then happens?”
  • Quote (D, 39:04): “There’s both a fine line as well as a yawning chasm between output and outcomes.”

D. The Rising Cost of Compute: Can Agencies Negotiate?

• GPU Costs Could Reshape Pricing:
  Agencies, like AI companies, incur massive compute costs; some may need to pass those costs to clients—potentially favoring usage-based compensation.

  • Quote (B, 40:29): “I imagine agencies would stand to also be losing money because of how astronomical these costs can be.”

• Bulk Bargaining Power:
  Like in media buys, agency holding companies may gain leverage by negotiating bulk rates for compute and AI tools—something smaller brands can’t easily do.

E. Implications for Jobs and Agency Model

• AI as Labor Replacement:
  Holding companies, through natural attrition and not replacing retired staff, are already saving significantly on headcount, though some savings are offset by AI costs.
  • Quote (D, 46:00): “CMOs at the holding companies are privately acknowledging that they're using natural attrition to just not replace people ... AI can kind of fill those gaps.”
• Harsh Entry-Level Realities:
  Entry-level roles most vulnerable to being eliminated or automated away, raising big questions for future talent pipelines—and possibly an explosion of freelance/independent shops enabled by AI.
  • Quote (D, 46:58): “Entry level jobs are just harder and harder to come by because that seems to be where a lot of AI can kind of be applied ... how do you become seasoned enough to be able to get that mid level job?”

F. Agencies Will Evolve, Not Vanish

• Holding Companies’ Endurance:
  Predictions of the agency holding company’s demise remain premature—even in an AI age, complexity and negotiating power keep them relevant.
  • Quote (D, 48:51): “All the death knells of agency holding companies ... there may be fewer in two years ... but I think agencies are going to always remain kind of essential to this process ... they always find a way.”
• A Changing, Not Dying, Profession:
  The coming years will have realignment, new roles, new models, and—perhaps most importantly—unanswered questions about what full-on agentic AI will do to the business of ads.
  • Quote (B, 51:19): “If we do move to an agentic web ... who are you really advertising to? At the end of the day you’re advertising to the AI models ... how much human intervention do you actually need?”

Notable Quotes & Memorable Moments

• “If you can’t beat them, sue them. But that really does seem to be this man’s philosophy…” (B, 05:50)
• “I think we’re kind of there with the AI bit now where you’re starting to see brands that have jumped on this thing early ... but then they’re starting to realize there’s a lot of work that needs to go into it.” (C, 24:07)
• “There’s both a fine line as well as a yawning chasm between output and outcomes.” (D, 39:04)
• “Entry-level jobs are just harder and harder to come by because that seems to be where a lot of AI can kind of be applied ... It seems a little bit like a catch 22.” (D, 46:58)
• “All the death knells of agency holding companies ... agencies are going to always remain kind of essential to this process because the process is complicated, the market is complicated.” (D, 48:51)
• “If we do move to an agentic web ... who are you really advertising to? At the end of the day you’re advertising to the AI models…” (B, 51:19)

Timestamps for Important Segments

| Segment | Timestamp | |-------------------------------------------------------|--------------| | Gap/American Eagle Ad Social Critique | 01:05-04:40 | | X.AI lawsuit & Apple/OpenAI AI drama | 04:40-11:22 | | Retail media bubble, Macy’s/Amazon deal | 11:22-15:40 | | AI publisher compensation, usage-based models | 15:40-20:19 | | FEATURE – Agencies, In-Housing and AI (Full Segment) | 21:05-51:12 | | - AI in-housing parallels, brands’ perspectives | 21:05-24:13 | | - Agency adaptation and business model changes | 27:26-32:53 | | - Agentic AI use cases, focus groups, measurement | 34:06-39:04 | | - Compute costs and agency negotiating power | 40:29-44:26 | | - Impact on jobs and entry-level talent | 46:00-47:27 | | - Holding companies’ future, summary note | 48:51-51:02 |

Conclusion

In this episode, Digiday’s team delivers a nuanced, up-to-the-minute look at how AI is a double-edged sword for agencies—simultaneously reducing costs, raising existential questions about agency value, upending compensation models, and threatening entry-level paths. Yet, amid all this turbulence, agencies’ adaptability, scale, and expertise suggest they are likely to endure—albeit in a rapidly evolving, digitally supercharged form.