Evertune: The Trade Desk Vets Building the Bloomberg Terminal for AI Search
Brian Stempeck scaled a $50B adtech company. Now he's betting GEO — Generative Engine Optimization — will become the most important marketing channel that no PMM is thinking about yet.
Brian Stempeck
CEO & Co-Founder, Evertune
The Origin Story
It started with a pool heater.
"I was connecting with former colleagues at The Trade Desk," Brian Stempeck told Pulse 2.0, "and we were discussing how often we were using AI platforms like ChatGPT, Gemini, and Claude to make purchasing decisions. At the time, I needed a new heater for my pool, so I went to these platforms to help me figure out what to buy."
What Stempeck and his co-founders — Poul Costinsky and Ed Chater, both Trade Desk alumni from the Adbrain acquisition — experienced was a revelation: these platforms are fundamentally reshaping how people discover and evaluate products. And there was no way for brands to measure or influence it.
"Today, 58% of consumers have replaced traditional search engines with generative AI tools for product and service recommendations. And 90% of B2B buyers already use gen AI to research vendors. This means that marketers need to understand how their brand is showing up in LLMs."
— Brian Stempeck, CEO of Evertune
The Team
This isn't a first-time founder pitch. The Evertune founding team spent a combined 20+ years scaling The Trade Desk from startup to $50B public company:
- Brian Stempeck — First commercial executive at The Trade Desk, rose to Chief Strategy Officer over 11 years. Has a board seat.
- Poul Costinsky — Co-founder of Adbrain, acquired by TTD in 2017. Deep experience in identity and cross-device targeting.
- Ed Chater — Also from Adbrain. Built data infrastructure at scale for programmatic advertising.
The investor list reads like an adtech hall of fame: Roger Ehrenberg (former TTD director), Rob Perdue (former COO), Susan Vobejda (former CMO), Jonah Goodhart (founder of Moat, sold to Oracle), Jay Friedman (CEO of Goodway Group), and Tal Chalozin (co-founder of Innovid).
Their $15M Series A was led by Felicis Ventures with participation from Eniac Ventures, NextView Ventures, and angels from OpenAI, Meta, and Uber.
The Problem They're Solving
Here's the uncomfortable truth: Gartner predicts 25% of traditional search volume will disappear by 2026. It's not going to another search engine — it's going to AI.
When your buyer asks ChatGPT "what's the best CDP for mid-market companies?" or Perplexity "which TV is best for gaming?", you have no visibility into whether your brand appears, how it's positioned, or what sources influenced the recommendation.
Traditional SEO doesn't help. These aren't search rankings — LLMs synthesize information from sources they trust and return a narrative recommendation. The same query asked 1,000 times will generate 1,000 different variations.
The Product: A Bloomberg Terminal for AI Search
Evertune is building what they call the "complete Generative Engine Optimization platform" — comprehensive tools to monitor, measure, and improve how your brand appears across ChatGPT, Claude, Gemini, Perplexity, AI Overviews, Meta AI, and DeepSeek.
🔍 AI Brand Monitoring
Track how often and how prominently your brand gets recommended by AI. Benchmark against competitors with statistically significant sample sizes.
🌐 AI Website Optimization
Audit how AI crawlers interpret your site. Many brands find their sites block LLM crawlers by default — Evertune identifies and fixes these gaps.
📊 AI Content Strategy
Decode which sources are most influential to each LLM. Get specific content recommendations and distribution targets to improve your AI Brand Score.
🎯 AI Retargeting
Programmatic ads on the URLs that AI cites most — via integrations with The Trade Desk and Index Exchange. No Evertune subscription required.
💡 Why Volume Matters
"LLMs don't give the same response twice. You and I could write similar prompts and get entirely different answers. That's why we analyze 1 million prompt responses monthly per customer — our competitors do 1% to 10% of that. Think of it like polling: would you ask 5,000 voters or 5?"
What They've Discovered
Evertune's research has surfaced counterintuitive insights about how LLMs work:
- 62% of ChatGPT responses come from pre-trained knowledge (the base model), not real-time search. You need to be embedded in the model's foundational knowledge, not just indexed.
- Each LLM has different source preferences. Gemini heavily weights Reddit and Quora (Google has data licensing deals). Llama overindexes Instagram and Facebook engagement. A shoe brand with poor Llama rankings traced it to Instagram complaints about arch support.
- Many influential sources block AI crawlers by default. If your glowing review on an industry blog isn't crawlable, LLMs will never see it.
Real Results
Evertune works with global brands including Canada Goose, Miro, and WPP's Choreograph. Here's what they're seeing:
📈 B2B Software: Top 10 AI Ranking in 2 Months
After competitive analysis revealed gaps in security and integration messaging, the client deployed 8 strategic blog posts, overhauled product pages, and secured media coverage on 3 high-authority publications. Result: top 10 ranking in AI recommendations within 60 days.
✈️ E-Commerce Travel: 15-Point AI Brand Score Increase
A/B test with AI-optimized content: the company created 30,000 words of new authoritative content using Evertune's messaging framework. Within 4 months, they saw a 15-point increase on ChatGPT vs. 3-point increase on Gemini and Meta AI (which were blocked from seeing the new content as a control).
🏢 SAP HCM: GEO Analysis Deep Dive
We ran SAP through Evertune's platform to evaluate AI visibility in the HCM category. See how SAP compares against Workday and Oracle across ChatGPT, Gemini, and Claude — including word association, sentiment analysis, and content strategy recommendations.
View SAP Analysis → (SAP access code required)
The Vision: AI as a Marketing Channel
Stempeck sees this as the beginning of something much bigger:
"We see AI as the most significant new marketing channel since social media. Think about how brands had to master Facebook ads, Instagram stories, and TikTok shop. Now imagine AI agents are booking restaurants, buying products, and making recommendations for millions of consumers every day. That's not a distant future — it's started happening."
— Brian Stempeck
The roadmap goes beyond monitoring. Evertune is building for:
- Agentic AI commerce — when AI agents don't just recommend products but buy them
- AI advertising — when LLM platforms start selling ad placements in conversations
- AI media activation — targeting the publishers and sources that most influence LLM recommendations
Why It Matters for PMMs
Traditional SEO was built for a world where humans searched, clicked, and browsed. That world is ending.
When an AI agent evaluates your product category, it doesn't click through 10 blue links, read your landing page hero copy, watch your explainer video, or download your whitepaper. It synthesizes information from sources it trusts and returns a recommendation.
If your brand isn't being cited — or worse, if it's being cited negatively — you're invisible in the fastest-growing discovery channel in B2B.
🎯 The Bottom Line
GEO isn't a nice-to-have anymore. It's table stakes for discoverability in an agent-mediated world. Evertune is early, but they're attacking a real problem with serious enterprise DNA. The question isn't whether this matters — it's how fast the shift happens.
What to Watch
Evertune just announced integrations with Index Exchange and The Trade Desk for AI Retargeting — letting any advertiser target the sources that most influence AI recommendations. That's the kind of product velocity you expect from people who've built category-defining infrastructure before.
For PMMs: this is worth understanding now, not later. Because by the time everyone's talking about GEO, the early movers will already have locked in their advantage.
Learn more: evertune.ai