Podcast Summary: AdExchanger Talks – “AI Won’t Shop For You – Yet”
Date: January 27, 2026
Host: Allison Schiff
Guest: Scott Howe, CEO of LiveRamp
Episode Overview
This episode explores the promises and real-world limits of artificial intelligence in advertising and commerce, with a focus on data privacy, industry upheaval, and the changing landscape for brands as AI and agentic decision-making reshape how advertising works. Allison Schiff welcomes Scott Howe—CEO of LiveRamp and self-described “king of collaboration”—to unpack the industry’s reset moment, how data companies are treading privacy minefields, whether AI agents will shop for you any time soon, and what makes or breaks data partnerships in a world where standardization collides with flexibility.
Key Discussion Points & Insights
1. Scott Howe’s Background and Perspective
- Fun Fact: Scott is a (tiny) part-owner of the Green Bay Packers (02:25), thanks to their public ownership model.
- His first startup, King of the Web, mixed the weirdness of social media and politics, but fizzled after initial buzz (05:02).
- Scott would choose to be “king of collaboration”—tying into LiveRamp's role as a data collaboration platform (06:39).
- “I have the best job in the world…because Liveramp is neutral and we work with so many folks…In the business of helping them out.” (07:02, Scott Howe)
2. What Does LiveRamp Actually Do?
- LiveRamp enables companies to make better decisions by making it easy to access, combine, and activate data—“the easy button for using data.” (07:46)
- Biggest value: interoperability—acting as “the Switzerland of data infrastructure.”
- Scott’s kids, part of the tech-native generation, “are more tech savvy than I am,” and one helps write California privacy bills (08:21).
3. The Privacy Minefield & Consumer Education
- The discussion covers how privacy legislation, lawsuits, and public confusion continue to challenge the industry.
- Scott’s Recommendations for Data Companies:
- Educate customers to reduce fear and make the value of data exchange clear.
- Get explicit consumer consent—make it “hard for someone to say, ‘Hey, I gave my permission, but I think it’s wrong.’” (10:25)
- “If you do those things…it’s really hard for someone to say, ‘Hey, I gave my permission, but I think it’s wrong.’” (11:52, Scott Howe)
- Ongoing challenge: Making data collection transparent and accessible remains unsolved despite decades of workshops and laws (12:19).
4. Embracing Change and AI’s Industry Reset
- Allison asks what Scott would tell his 2019 self; his advice: “Move faster!” — avoid errors of inaction, especially in a rapidly evolving industry (14:09).
- Example: In 2015, dominant DSPs (Turn, Rocket Fuel, AppNexus) don’t exist today; change is constant.
- “Lean in, embrace the change, stay a few steps ahead…you don’t need to be miles ahead…just a few steps.” (15:38, Scott Howe)
5. “Spicy” AI Predictions & Agentic Commerce
- Scott is skeptical of rapid adoption of AI agents that shop for you:
- “We’re using AI to collect information, not necessarily perform the action right now.” (18:05)
- Cites Perplexity’s shift in ad strategy: pairing user memory with advertiser inventory for more useful responses, reimagining what ads can be (18:05–21:25).
- Focus should be on “moments of despair and moments of delight.” Use AI to solve for customer frustrations and create delightful, personalized experiences (20:30).
- “If we can harness the power of AI on just those two things…every company can figure out some really interesting things they can do with AI.” (21:20, Scott Howe)
- Netflix’s recommendation algorithm is held up as a positive case study (20:30).
6. OpenAI’s ChatGPT Enters Ads: Relevance and Integrity
- OpenAI’s plan to add ads in ChatGPT’s free tier—claims that “answers are never driven by advertising” meet with skepticism.
- “I visited Amazon in ‘99…they got really upset and said, ‘We will never do advertising.’ Lo and behold, here we are…one of the biggest advertising companies in the world.” (29:06, Scott Howe)
- Scott’s stance: As long as ads improve the user experience, they’re legitimate. The computing costs of AI make advertising a “logical choice.” (29:06)
- Discusses challenges: ads must be relevant and well-integrated, or they become ignorable “noise” (30:55).
- Allison describes being (surprisingly) effectively targeted by a simple display ad—proof that relevance can break through (32:49).
- References Sam Altman admitting Instagram is an example of good, relevant advertising (33:45).
7. AI Search & The Battle to Stay Present in Answers
- AI search has doubled in usage but is still a fraction of traditional search; the nature of “zero click” search is changing both user habits and marketer strategies (34:35).
- Brands now compete to be present in AI-generated answers, not just inventory spaces.
- LiveRamp’s response: Build tools for clients to activate their data into AI-powered channels—now partnering with 17 AI providers (Google, Meta, Perplexity, others) (36:04).
- “Winners and losers haven’t yet been determined…We will work with all of them.” (39:40, Scott Howe)
8. Making Data “Agent Ready” & The Need for Standards
- Data must be structured, governed, and interoperable to feed into AI agents—governance and visibility are “absolutely critically important” (40:33).
- Centralizing permissions and control for partner access is a role LiveRamp wants to provide (41:41).
- Standards are proliferating—“It does feel like we need standards for our standards here.” (42:31)
- Scott: “It doesn’t matter who builds the standards…[as long as] it’s democratized and everybody’s using it” (43:13). Interoperability unlocks value; proprietary approaches shut down collaboration.
9. The Holding Company Shakeup, Standardization vs. Flexibility
- Ongoing holding company mergers (Axiom at Omnicom, IPG, etc.) are reshaping identity and data practices (44:37).
- Scott’s “mantra” has shifted from “standardization and scalability” to “flexibility”—partners all have unique needs and recipes for success (45:20).
- Goal: LiveRamp as the “Intel Inside” of data, powering every major data-driven company or platform (47:10).
- On M&A rumors (Publicis acquisition): “I love the fact that people are talking about us, and I will never, ever comment on anything specifically.” (48:26, Scott Howe)
10. Buzzwords That Should Die
- The one ad mantra Scott wants to kill: “The customer is always right.”
- “It’s just wrong.…Recognize their pain is real, but their solution may not be the right approach…Solve the problem, not just do what is asked. Anybody can be an order taker.” (49:33, Scott Howe)
- Allison: “Sometimes the customer is just wrong.” (50:51)
Notable Quotes & Memorable Moments
-
On AI changing advertising:
“The moments that matter are moments of despair and moments of delight. If you can harness the power of AI on just those two things, then almost every company can figure out some really interesting things they can do with AI.”
(21:20, Scott Howe) -
On moving fast in ad tech:
“Lean in, embrace the change, stay a few steps ahead…you don’t need to be miles ahead…just a few steps.”
(15:38, Scott Howe) -
On AI agents:
“We’re using AI to collect information, not necessarily perform the action right now.”
(18:05, Scott Howe) -
On advertising’s evolution:
“As long as you always have the guiding principle of we’re going to do what makes sense for our users and deliver a better customer experience…that may include advertising, it may not include advertising.”
(29:06, Scott Howe) -
On data standards:
“We don’t care what standard is embraced. We just want to make sure that whatever the standard is, it’s democratized and everybody in the industry is using it.”
(43:13, Scott Howe) -
On buzzwords:
“The mantra of the customer is always right. It’s just wrong.…Recognize their pain is real, but…solve the problem, not just do what is asked.”
(49:33, Scott Howe)
Timestamps for Key Segments
| Segment | Topic | Timestamp | |---|---|---| | Scott’s Packers story & King of the Web | Fun intro, background | 02:25–06:39 | | What is LiveRamp? | Jargon-free company summary | 07:46–08:12 | | Data privacy & lawsuits | Industry challenges | 08:54–12:43 | | Advice to 2019 self & change in ad tech | Move faster lesson | 14:09–16:12 | | “Spiciest” AI takes & agentic commerce | AI’s true potential | 17:13–21:25 | | OpenAI, ads, and user experience | ChatGPT ad rollout | 27:48–34:35 | | AI search, changing consumer journey | From search to answers | 34:35–40:33 | | Making data agent ready | Data governance/standards | 40:33–44:37 | | Holding company drama & LiveRamp’s role | Standardization→flexibility | 44:37–48:21 | | Buzzwords Scott hates | “Customer is always right” | 49:33–50:51 |
Summary Takeaways
- AI is changing everything, but not as quickly as “agentic commerce” evangelists predict—most people use AI to research, not actually transact (yet).
- Making data useful for AI requires more than just access: structure, governance, and clear value exchange are non-negotiable.
- Flexibility rules: Data solutions must adapt to unique client recipes, standards, and constant ecosystem shakeups—even as they press for interoperability.
- Ads will creep into every platform where there is attention—eventually in places like ChatGPT—driven by utility and economics.
- The next phase is not about miles-ahead innovation, but nimble adaptation to unknown “winners” in AI and media.
For anyone in ad tech, marketing, or digital media, this episode delivers a rich, realistic look at the tectonic shifts ahead for data, privacy, and the continuing (and messy) evolution of advertising in the AI era.
