Welcome to ThDPTh: Where data & AI hype goes to die
Join me to become the person who spots the data & AI BS before your team wastes six months on it.
Hello,
I’m Sven. I was Head of Marketing at a data startup and led the founding team through three pivots before we figured out what actually mattered. I was a PM at a data & ML platform serving 700 people where I watched brilliant engineers build tools nobody used. (We did have a couple of big successes, too ;))
Now I’m building amazing AI features at MAIA, where “AI is our tool, not our purpose,” and coaching AI, data & open source startup founding teams on the side.
I’ve spent 12+ years in the BI, data, and AI industry (though it was still “ML” when I started out), and here’s what I’ve learned:
Most advice about data, AI, and business intelligence is bullshit.
Not intentionally—but the gap between vendor promises, LinkedIn heros and implementation reality is enormous. The space between “data-driven company” marketing and what actually drives better decisions is a canyon.
This newsletter exists to close that gap.
About Data & AI
Let’s make one thing clear from the beginning: You’ll get both data and AI stuff here, because I have learned, experienced one single lesson that has burned itself into my brain: “it’s just math.” (as a PhD in mathematics, I don’t say that lightly)
Becoming a data driven company? Creating amazing AI-powered products? It all comes down to the same fundamental lessons, rooted in data, in digitized abstractions of physical events, to essential math (itself a beautiful abstraction, and invention of us humans). It’s not hype, it’s natural.
What You’ll Get Here
Contrarian attacks on industry consensus — Like “What’s Wrong with BI“ where I explain why BI tools are built for the wrong users, work with the wrong data, and use the wrong interface.
Case studies from companies that actually pulled it off — How Airbnb became data-driven, what Zynga understood that others missed, and the specific playbooks that work at scale.
Frameworks you can actually use — Not listicles. Not hot takes. Reusable mental models that help you make better decisions. Things like “Escaping Data Flatland” or understanding why dashboards suck (and how to make them suck less).
What’s really happening with AI — Not the hype. Not the fear. What’s actually working when you implement it. I’m currently walking the talk at MAIA, building AI-powered tools where the product principle is clear: AI is a tool, not the purpose. That perspective shapes everything I write about AI adoption and implementation.
Who This Is For
Based on the 2,000 people already here, you’re probably:
Data & Analytics Engineers (30%) — Building dbt pipelines, working with Snowflake/Databricks, tired of dashboards nobody uses
Product & Strategy folks (24%) — Data & AI PMs, Product Leads, Heads of Data/Analytics trying to figure out what “data-driven” actually means
Leaders & Founders (20%) — CEOs, CTOs, CDOs making platform decisions with real money on the line
Software & AI Engineers (12%) — Building AI/ML systems, full-stack developers, watching the gap between AI hype and reality
BI/Analytics professionals (9%) — BI Engineers, Analysts, Data Scientists who know the current approach isn’t working
GTM folks (5%) — Product marketers, growth people, RevOps trying to use data that actually drives decisions
You’re technical enough to know when something is BS, senior enough to care about what actually works, and honest enough to admit when your current approach isn’t landing.
What This Isn’t
I don’t do:
Tool reviews (I’ll tell you why certain approaches fail, but I’m not your curator)
Trend-chasing hot takes
“13 lessons I learned” compilations
Anything I haven’t personally built, broken, or fixed
If I write about it, it’s because I’ve implemented it, watched it fail, or had real money riding on the decision.
That said: When I do publish something, I put in a TON of research. I read 2 books a week on top of everything else. So when you read a piece here, know that it’s deeply researched—I just don’t write about things until I’ve actually done them.
Why ThDPTh?
2,000 of you are already here. Data engineers at Snowflake and Databricks shops. Product leaders at B2B SaaS companies. CTOs and CDOs making platform decisions. AI engineers implementing LLMs in production.
The common thread? You’re tired of the BS. You want to know what actually works when the consultants leave and you’re stuck with the implementation.
That’s what this is.
One More Thing
This newsletter is free. No paywalls. I write this because I think the industry needs more honest conversation about what actually works versus what the slide decks promise.
If you find it valuable, share it with someone who’s fighting the same battles. That’s how this grows.
Welcome aboard.
— Sven
P.S. — Got a dashboard nobody looks at? A data strategy that’s going nowhere? A tool decision you’re regretting? Hit reply. I actually read these, and half my best articles come from conversations with readers & leaders who are in the trenches.
2,000 subscribers and I've never written an intro post.
Why? I've been too busy explaining why your BI tools don't work, why dashboards sit unused, and what companies like Airbnb actually did to become data-driven (versus what the case studies claim).
But hitting 2,000 felt like the right time to finally explain what ThDPTh is—and who it's for.
Grateful for everyone here, especially the readers who share their own implementation horror stories. That feedback is what keeps this real.