insights >>
Video
June 4, 2026

Why Most 'Data-Driven' Companies Are Fooling Themselves | Alex Lirtsman, CorralData

Proof of Work — Episode 097
Why Most AI Spending Never Reaches the P&L
Jimmy Bijlani, AI Momentum Partners
00:00
--:--

Alex Lirtsman, founder & CEO of CorralData, joins Stu Willson to make the case that most "data-driven" companies are really just "data-adjacent." After ~15 years building and selling the digital agency Ready Set Rocket (clients including the NBA, Pfizer, and Sweetgreen), Alex started CorralData to automate the entire data stack — unifying 600+ fragmented data sources, generating reporting, and now letting AI agents take autonomous action on insights, all within HIPAA-compliant boundaries. The conversation covers why appointments, leads, and ROAS are vanity metrics; what it actually takes to bridge EMR, CRM, and paid-media data to get a real LTV:CAC picture; the move from read-only AI to autonomous execution (and why "just because you can act doesn't mean you should"); a detailed case study from a PE-backed med-spa platform; and why brand and data are the only two assets a company can't rebuild.

Keywords: Stu Willson, Stuart Willson, Just Curious, Pluris, Alex Lirtsman, CorralData, data analytics, AI agents, autonomous AI, data-driven, data-adjacent, healthcare data, HIPAA compliance, LTV CAC, marketing attribution, private equity, med spa, operational effectiveness, business intelligence, Reverse ETL, data warehouse, AskCorral Agent

chapters
19 marks
00:00
Intro: who is Alex Lirtsman and what is CorralData
01:07
The one thing he wants you to do: just get started
01:47
From digital agency to automating data ("Shopify for data")
02:47
Why he turned 15 years of consulting into a platform
04:06
Why every report and dashboard has a dead end
05:33
What changed — technologically and culturally — to make this possible
07:41
How CorralData actually works: connect, warehouse, semantic layer
10:32
Beyond dashboards: talking to your data and acting on it
12:26
The hidden value in EMR/ERP data — and bridging systems
14:55
Data-adjacent vs. data-driven
17:23
What teaching at Columbia/NYU taught him about usable data
18:54
From read-only AI to autonomous execution
19:55
Guardrails: just because you can doesn't mean you should
22:10
Where leaders go wrong when they say "we need AI"
25:27
Case study: a PE-backed med-spa optimizing appointments vs. revenue
29:15
How long it takes to unify EMR + CRM + paid media
32:47
What changed first once the data was connected
34:40
ROI: customers growing ~7x faster
Working on something like this?
Submit a need and practitioners respond with how they would approach it. Free, anonymized.
submit a need

More insights

Sep 29, 2026
Someone Already Solved the AI Problem You're Stuck On
Stu Willson
Sep 22, 2026
Assessing AI, when you cannot see inside it
Dhruv Gulati
Sep 15, 2026
The Migration Can Wait
Jordan Gurrieri