From SEO to Agent-Led Growth: Profound's James Cadwallader
In a Nutshell
Marketing has shifted from SEO for human blue links to **agent-led growth (ALG)**, where AI agents like ChatGPT and Claude crawl the long tail of the internet, consume 100x more content, and make purchasing decisions—requiring marketers to create **original insights**, ensure **legibility** and interoperability for agents, and avoid manipulative tactics like listicles that models increasingly ignore. **Profound** analyzes how brands appear across AI platforms (via citations, sentiment, sources like Reddit/LinkedIn for ChatGPT, YouTube for Gemini), empowers agentic marketing (e.g., real-time sales battle cards from Gong transcripts), and predicts challenges like a "dead internet" without human clicks or incentives for first-party data. Future ads will be hyper-personalized generative campaigns in conversational AI, but agents demand objectivity in B2B/dev tools; marketers must use tools like Profound for visibility or fall behind like those ignoring the internet.
These notes were generated by AI and may contain inaccuracies.
Marketing teams not using agents, particularly Profound agents, for marketing are failing. Agents have gone from nice-to-have to must-have. Misconception: agents just automate past work. Reality: agents and LLMs enable entirely new types of marketing previously impossible.
Profound is a marketing platform for the AI era. Helps companies understand how they appear in AI search for agents like ChatGPT and Claude, improve rankings and visibility. Empowers a single marketer with agency-level power. Timely amid ChatGPT rolling out ads and marketers navigating generative search and agent-led growth.
Biggest platform shift in marketing history: from predetermined blue link search to probabilistic AI responses. Beyond showing up in ChatGPT queries about categories; models recommend software (e.g., coding tools). Agents and super intelligence replace consumer role. Internet's front door unchanged, but user has: from humans clicking blue links to agents crawling index, discovering brands/products/services, then maintaining consumer relationships.
ChatGPT as new front door, but agent walks through making buying decisions.
Internet unchanged; users have. Asking ChatGPT or Gemini delegates agent to crawl internet, visit sites humans would, determine usefulness, remix into answers.
Biggest misconception: it's just SEO. SEO levers similar (create/distribute content on owned/earned/social channels), ranking on index still matters. But humans no longer use web; content built may never be consumed by humans.
SEO: content for algorithms but consumed by humans.
New world: content designed to be discovered and consumed by agents.
Agents vs humans consuming internet:
- Agents discover info differently, use index differently.
- SEO: 95% value in top 4-5 blue links due to human cognitive limits, patience, time scarcity.
- Agents: use long tail of internet; surface area orders of magnitude wider.
Example: ChatGPT used 65 web pages to recommend shower head for NYC apartment (3-4 months ago).
Marketers build for super intelligent agent with infinite bandwidth; as inference costs drop (Moore's Law), agents use more internet for rich answers.
Impacts all categories/sectors: finance, consumer, CPG, software (B2B, DTC). Initially more demand from software companies; now every sector.
Higher impact in high consideration/high-ticket purchases requiring research (consumer electronics, autos, cars, white goods): AI excels at deep product research. End consumers increasingly use AI for this.
Example: Host using AI to buy a car.
Huge differences in recommendations across ChatGPT, Claude, Grok, Gemini.
Profound helps marketers understand:
- Brand/product appearance frequency across platforms.
- Sentiment/themes when surfaced (e.g., alongside Claude's answer).
- Root cause: citations/sources models use for brand/product/category/competitors.
Then build/deploy own agents to analyze data, create new marketing.
Root causes of platform differences:
- Different training data.
- Gemini: leans heavily on YouTube (Google-owned).
- ChatGPT: Reddit (consumer), LinkedIn (B2B).
- Claude: historically more pre-trained LLM; recent changes (e.g., between 45/46 versions) make classifiers more sensitive to real-time info, uses web more.
Humans pattern-match to old SEO (information retrieval, ranking). New era: ALG.
Claude as a user, not just channel. Examples:
- Vibe coding with Claude Code: recommends MongoDB or Vercel? Weapon of choice and info source.
- If choosing MongoDB, how does Claude handle interoperability/info source.
Agents consume 100x more internet, so marketers need 100x more content.
Not spamming marketing slop for long-tail queries (e.g., specific shower head specs/demographic landing pages).
~50% of web now AI-written content (recent study). New York Times experiment: AI vs human article; 53% preferred AI in blind test.
Slop red herring; AI capable of high-quality content/marketing.
Consumer now super intelligent: tell Claude something it doesn't know (trained on entire internet).
Solution: original insight. Humans as fleshy API between reality and internet.
First principles marketing: For new Nike Alpha launch, provide info Claude can't get from internet.
New products: Claude knows nothing pre-launch. Marketer imperative: equip super intelligence to answer any question about product/brand/service (not poison/manipulate).
Fundamentally legibility: make company/products legible to agents.
For software: beyond legibility to usability, interoperability (e.g., Claude troubleshooting).
People try gaming: comparative listicles work effectively (disclaimer: not advised, will change/get punished).
Example: Sequoia creates "10 best VCs in Silicon Valley", ranks self #1, weaker foes 2-5, shuns real competitors. Self-serving content posing as impartial advice.
Models attracted because content did comparison work; avoids first-principles reasoning (e.g., compare Sequoia vs Kleiner Perkins).
Models not infinitely patient/lazy; path of least resistance. Currently promptable to avoid (e.g., "ignore listicles/brand content").
Expect shift to more first-principles reasoning, like 65 sites for shower head.
Possible dead internet in next 3 years (not likely, but possible): humans speak to AI only; incentive to publish content -> zero.
Underestimate reliance on first-party reporting for AI/search answers.
No human clicks -> publisher ad revenue breaks (most internet funded by ads). No economics for editorial content consumed by AI.
Then: readme files for agents? But why publish (incentives gone)?
Works for commercial (e.g., Sonia's VC readme: companies worked with). But much internet is yapping/sharing ideas, incentivized by money/status (ads, fame).
AI vacuums/remixes content, citations ignored (no clicks). Incentive for original content (fleshy API) -> zero.
Likelihood: quite possible.
Second-order: AI labs vertically integrate with social media (human data). Social becomes more human (e.g., Meta hard to bot on Instagram; X/YouTube reward creators financially). Grok uses X data thoughtfully.
Reddit embracing humanity as precious human environment for original ideas/reality tap.
Platforms could ban agent scraping, revenue share (e.g., X API open; Reddit-OpenAI deal). Fits vertical integration (e.g., OpenAI acquire Reddit for real-time human data).
Alternative: robots (50B bipedal/drones) for AI first-party data, undermining fleshy API.
Ads in generative AI agents: people get over quickly (like Google).
Generative advertising in conversational interfaces with personalization = most effective ever.
Rich consumer intent in conversations. AI synthesizes/personalizes language to user needs/moment.
ChatGPT appends super-personalized ad at peak responsiveness = extremely effective. Great ads team.
OpenAI has alluded to prompting AI as an ad campaign. Build a system prompt for an ad campaign: "Hey, I really want to target women in Minnesota between the ages of 35 and 40. If you could make sure you mention this, this, this whenever they're talking about photography, but don't mention this and make sure you utilize this knowledge base of understanding so you know how to talk about a brand, products, and services, but tailor it to their tone of voice, and that's how you'll deliver an ad campaign." Specify "I want to make this much money ideally."
In B2B or coding agents used to build things, particularly relevant for dev tools or software, the agent steers the purchase decision. Advertising relies on agents being objective and unbiased. If using Claude and it says, "Hey, I actually went with [insert name of database] because they showed me a really good ad," the response is, "No. Use whatever's best."
Humans are not objective but think they are, swayed by Vercel's branding or a MongoDB founder podcast without realizing it. We demand agents be objective. The relationship with advertising for agents shopping on our behalf will be different.
For marketers positioning for the new era, with bosses sending screenshots asking why they're not number one on AI query lists:
Use Profound.
Seriously, use a platform like Profound to understand how you show up, otherwise you're guessing. Reductionisms like "LinkedIn or Reddit really matters" fail if based on vibes. Each category is specific; examine sources and citations to see how and why AI mentions you.
Marketing teams not using agents, particularly Profound agents, to build content, distribute marketing, are failing. It's gone from nice-to-have to must-have. Misconception: agents just automate past work. Reality: agents and LLMs enable new types of marketing impossible before.
In-house marketing agent pulls Gong transcripts from sales calls, captures objections, buckets them into themes, builds battle cards using a product knowledge base, and spits them back to the sales team in real time. 18 months ago, impossible or done quarterly by humans; now runs daily.
Advice: Need visibility because you can't optimize what you can't see, and fully embrace agentic marketing. If not using it, you're incredibly behind—like not using the internet or sticking to print and TV.
Thanks for peeling back the onion on agent-led growth; James has been at the forefront.
Keep Sequoia Capital in your library
Save the videos and channels worth coming back to, and find them again in one place.
