Sam Tomlinson has two full-time employees whose only job is figuring out how to use AI to market better. He's audited nearly 100 ad accounts a year, runs a marketing agency of 50 people, and operates a venture capital fund investing in tech.
At SLIF, he skipped the hype and got straight to what works – and what doesn’t.
A leading authority on the intersection of performance marketing, digital advertising, analytics and technology, his philosophy is simple: "The cost of answers goes to zero. The cost of creation goes to zero. But the value of asking the right questions and applying AI to the right things? That's infinite."
Here's the inside baseball of what his team has learned.
Every operator says they know their prospective clients. But when was that research actually done? Once. Maybe twice. And it focused on the wrong things.
Tomlinson uses SparkToro, which pulls data from 120 million US profiles showing exactly what people search, which subreddits they visit, what YouTube videos they watch, and which influencers they trust. Upload that into a custom ChatGPT persona builder and you get a complete picture of how that prospective client actually uses the internet today, broken into research, evaluation, and decision phases, each with specific pain points and emotional triggers.
When his team mapped those insights onto five core human motivators (survival, admiration, success, love, and protection), they built an automated creative workflow that reduced cost per lead by up to 50%, doubled lead volume, and increased qualification rates by 130%.
With AI’s lightning fast march toward ubiquity in the US, time is of the essence.
Tomlinson's team pulled 125 AI overviews in senior living, collected all the content and links inside them, and ran everything through a pattern-finding analysis to figure out what they all had in common. Essentially, they worked backwards from the results to find the formula. Eighty percent of inclusion comes down to just four things, they found:
Exceptional content (Google's definition: well-researched, in-depth, accurate, helps people accomplish tasks. Not 500-word blogs).
Written for humans. Tomlinson's test: read your article out loud. If you want to jump out a window halfway through, rewrite it.
Multiple media formats. Most senior living blogs are just text. Articles that get included feature infographics, videos, Q&As, social posts.
Brand mentions across trusted sources. Large Language Models don't just count links. The more your brand is mentioned alongside your topic, the more likely you are to appear.
Four content types that AI overviews favor: guides and checklists, links to authoritative nonprofit or government sites, mentions on industry websites, and resource hubs.
Why does this matter? Google permanently added ads to AI overviews in late 2025 and has been rolling them out across AI mode ever since. Currently 53% of searches serve an AI overview, and they're undermonetized. That's changing fast.
Tomlinson's team built an automated system that looks into competitor communities weekly: pricing, specials, which units they're advertising, how they're messaging. It exports everything to a Google Sheet, analyzes the patterns, and sends a weekly text summarizing every change.
The result: real-time intelligence that lets marketing teams react instantly. If a competitor shifts promos to memory care, pivot to market assisted living. If they fill up on memory care and stop advertising it, push harder in that direction.
Tomlinson ended with a rapid-fire list of practical AI applications his team uses daily:
Handwritten notes. When a qualified lead submits a form, ChatGPT generates a personalized note that gets printed and physically mailed to their home for $14. "In a super emotional, personal decision, somebody filled out a form and got a handwritten card. All from AI running in the background."
Sales call analysis. All sales call recordings are uploaded via API to Gemini, which evaluates every call, identifies what behaviors drive conversions, and flags the best ways to handle common objections. "I don't have time to listen to 10,000 phone calls. Gemini does."
Lead scoring. Tomlinson's team had AI run through an entire client database, identified the six biggest predictors of move-ins, and built a lead scoring matrix that hit 80% accuracy.
His message isn’t about transforming everything overnight. It was about finding the small, contained applications that compound over time. "Don't try to boil the ocean. Just make a cup of tea. The little things matter a lot more than you think.”
Watch the full talk below...