Podcast Summary: "Marketing in the Age of AI Answers"

AdExchanger Talks | Host: Allison Schiff | Guest: Alex Sherman (Co-founder, Bluefish, former CEO, Promote IQ)
Date: September 23, 2025

Episode Overview

This episode explores the profound changes AI-powered search and generative AI are bringing to the world of marketing. Allison Schiff interviews Alex Sherman, co-founder of Bluefish—a startup designed to help brands optimize and monitor their presence in AI chat and search results. The discussion delves into the origins of retail media, Sherman's journey from programmatic ad tech to AI, the specifics of AI search optimization, current industry challenges, and the future of AI-driven marketing channels.

Key Discussion Points & Insights

1. Alex Sherman's Background and Early Career

• Fun Fact: Alex reveals he's half French, often surprising colleagues with his language skills during business meetings in Paris.
  [02:37]

  "I'm actually Half French. ... They see me as the American, ... but then I will trot out a bit of French as my party trick."

• Academic Roots: Both Alex and Allison have history degrees, which Alex credits for broad exposure and adaptability.
  [05:34]

  "One of the best parts of being a history major is you get to just sign up to learn about history all over the world... It gives you sort of a broad base of exposure."

• Entry into Ad Tech: Post-2008 financial crisis, Alex accidentally enters ad tech, starting at MediaMath—a "true meritocratic work environment."
  [09:59]

  "You tell people, 'Oh, I work at a startup,' ... everyone's like, 'Oh, I'm so sorry, like something horrible has happened to you.'"

2. The Birth and Evolution of Retail Media

• Founding Spotfront/Promote IQ: Initially, Alex and team aimed to democratize programmatic tech for SMBs, but pivoted after interest from retailers seeking control of their own ad networks.
  [14:36]

  "Every retailer had basically signed up for Hook Logic... But they had no idea what was getting pumped through that pipe... So in the early days, our pitch was: 'you’re not getting your fair share' and 'you should control this.'"

• Early Market Skepticism: In 2014, building ad tech for retailers was widely dismissed as a bad idea.
  [15:08]

  "In those days... you could not have possibly had a more unpopular business than an ad tech business."

3. Acquisition by Microsoft and Industry Shifts

• Life Inside Microsoft: Promote IQ went from startup "pirate ship" to scaling retail media within a corporate giant.
  [17:11]

  "Within Microsoft, if you build a billion-dollar business, you are a small division... It was a dramatic transition—everything just has to happen at 1000x the scale."

• Microsoft Pulls Back on Ad Tech: Microsoft later shut down their own onsite retail media offering, shifting to partners like Criteo. Sherman sees this as difficulty keeping pace and a signal that "AI is a little bit of a reset button."
  [21:02]

  "If you’re building a DSP right now, probably not the right time... Retail media feels very end-of-cycle... everything will exist differently when ads are embedded within agentic conversations."

4. The Rise of Generative AI Search (GenAI) and Bluefish's Solution

• Changing Consumer Habits: Both Allison and Alex have largely transitioned from classic search engines to AI-driven search.
  [24:01]

  "There’s a new expectation of an instantaneous, synthesized answer."

• How Bluefish Works: Brands specify which products, segments, and criteria to monitor. Bluefish generates synthetic prompts, syndicates to major AI platforms daily, and analyzes semantic output to assess brand visibility, sentiment, and competitive positioning.
  [25:03]

  "We built a marketing platform for AI... Brands designate what they want to track... Our engine takes those inputs and generates a large list of synthetic prompts..."

• High-Volume Analytics: To achieve reliable insights, Bluefish runs millions of prompts, using granular queries to accurately reflect real consumer journeys, and addresses the variance and personalization in LLM answers.
  [29:34]

  "The more prompts you track, the more resolution you get... The net result is a high volume of prompts for sure."

5. GEO: Generative Engine Optimization vs. Classic SEO

• Beyond SEO: Alex argues "GEO" is the next evolution—touching not just search, but content, PR, paid editorial, and more. The entire martech stack will need to adapt.
  [31:46]

  "...This goes way beyond search... Content teams, PR, corp comms—now they have a new vector to track."

6. AI Model Vulnerabilities & Industry Speculation

• Prompt Flooding, Quality Control, & Gaming Models: Current models may overvalue non-brand, sometimes low-quality content, incentivizing improper optimization and possible manipulation—issues providers will have to address for trust and quality.
  [32:15]

  "Right now, the models are learning more from content that didn't come from the brands than content that does... That might sound like a good thing, but it actually creates a perverse incentive... to trick the models..."

• Preventing Bad Practices: Sherman anticipates both brands and AI providers will need to adapt policies to prevent malicious manipulation, much like the maturing of the classic search/SEO industry cycle.
  [45:35]

  "It’s like the early days of search... I really do think the AI providers... have an incentive to create the best possible product for consumers. ...There will be misbehaviour—absolutely. But...history rhymes...we will get to a more stable, mature place..."

7. The Limits and Risks of Generative AI

• Inconsistency of Results: The same prompt may yield different responses by design. Bluefish solves this by large-scale, multi-profile querying and segment-level analysis.
  [37:09]

  "AI and GEO is really easy to do badly and really difficult to do well... That is why we take a high-volume approach..."

• Hallucination Risks: Some model hallucinations are wild ("eat a rock daily"), but more insidious are plausible-but-false subtle inaccuracies (e.g., CEO mis-naming, fabricated product details).
  [49:18]

  "The real issue is... subtle gray hallucinations where it is credible, it’s just wrong... because it’s so believable, it's even more problematic."

Notable Quotes & Memorable Moments

• On Startup Stigma
  Alex Sherman, [09:59]:

  "In 2008, if you're at a cocktail party and you tell people, 'Oh, I work at a startup,' ... everyone's like, 'Oh, I'm so sorry, like something horrible has happened to you.'"

• On Early Retail Media Skepticism
  Alex Sherman, [15:08]:

  "In those days ... building an ad tech business was absolute kryptonite. People would throw rocks at you. ... [But] every retailer that joined our platform ... it just started accelerating on its own."

• On AI Resetting the Playing Field
  Alex Sherman, [21:00]:

  "AI is a little bit of a reset button for ... [adtech]. If you’re building a DSP right now, probably not the right time..."

• On Evolving Search Expectations
  Alex Sherman, [24:01]:

  "There's ... a new expectation of an instantaneous synthesized answer. And I don't think I'm alone in that experience."

• On the Challenge of GEO
  Alex Sherman, [37:09]:

  “AI marketing in general is … really easy to do badly and really difficult to do well.”

• On Model Manipulation
  Alex Sherman, [32:15]:

  "There’s lots of really low-quality content ... the models are learning from, and there’s relatively less from the brands themselves. ... I wouldn’t be surprised if the AI companies try to move that slider to a more balanced position."

• On Marketing Channel Cycles
  Alex Sherman, [45:35]:

  "...history rhymes... there is a pattern, there is a cycle that plays out every single time ... we’re in the early stages of the next super cycle [and] AI is the channel that is eating all the other channels."

• On Hallucinations
  Alex Sherman, [50:13]:

  "My all-time favorite is definitely the Google Pebbles one ... advising users at least one rock per day."
  Allison Schiff, [52:15]:
  "What are you going to wash down your rocks with?"

Timestamps for Key Segments

• [02:37] Alex’s French background, icebreakers, and speaking French in business
• [08:03] Early career path & joining MediaMath in 2008 post-crisis
• [12:15 – 16:45] Founding Spotfront/Promote IQ & the origins of retail media
• [17:11 – 21:56] Microsoft acquisition, corporate transition, and strategic retreat from direct ad tech
• [23:02 – 26:23] Adoption of AI search in daily life, introduction to Bluefish
• [25:03 – 29:34] How Bluefish operates: prompts, LLM analysis, and high-volume monitoring
• [37:09 – 40:15] AI inconsistency, need for scalable measurement, and analysis strategy
• [41:13 – 44:56] How brands should respond to negative/variable AI-generated brand outcomes
• [45:35 – 48:09] Risks of negative competitive tactics, ad cycles, and future AI-provider responsibility
• [49:18 – 52:15] Hallucinations: from minor inaccuracies to infamous "eat a rock" moments

Conclusion

This episode delivers a candid, sometimes humorous, and always insightful look at the radical transformation underway in digital marketing due to generative AI. Alex Sherman recounts lessons from building ad tech during the rise of retail media, draws parallels to the new AI era, and explains the nuanced technical and ethical challenges that brands face as they try to control their narratives within AI-generated answers. Listeners get practical context—grounded both in history and in prognostication—for understanding what’s happening now and what’s coming next in marketing and search.