People generally think of projects sequentially (i.e. like gantt charts). However, data science is inherently circular
If you’re building an AI product. Don’t start with AI. Why? Users don’t automatically trust AI.
“When it came to free throws, I was willing to wait until the shooters started missing
Sales gets a bad rap in Product circles. I think they deserve our respect. Delivering revenue
Want to save yourself time and disappointment? I was getting fooled by false positives, so I
I’m grateful for entrepreneurs. Grateful for their vision. Grateful for their sacrifice. Grateful for their grit
There’s a “first time tax” for every new data set, problem area, and modeling technique. Great
Do your product strategies start out too complicated? Mine do. 🙋♂️ Think of complexity like Alcatraz.
Data features can easily become “underwater” Pay attention to Feature Economics. Impact = Value Created –
We’d laid off half the company 6 months prior. Only 8 of us remained. We were
Product Management changes dramatically as startups scale. Here are 5 ‘product stages’ startups navigate as they
Occam’s Razor. The longer you build data products, the more you realize how true (and important)
Here’s the article that goes with the video.
Averages lie…especially when evaluating AI accuracy. Is an 80% model accuracy good? Here’s how you know:
Prospects don’t like to change. These are my two favorite Product Messaging frameworks for catalyzing change:
Gone are the days of simply ‘doing AI’. Leaders expect AI strategies leading to real value.
Joshua Schultz and I talk about AI Product Management and lessons for SMBs on Racket.com: – How AI
Machine Learning projects often die, because of poor feature economics. Why? ML is expensive to build,
4 ways to know (2 obvious, 2 less obvious) 1. Accuracy -> Predicted vs. Actual Compare
Users had highlighted MULTIPLE feature gaps. After months of effort, I felt defeated. Then one of
Andy Grove famously claimed: “Only the Paranoid Survive” Healthy Paranoia can save your AI project. Here
Job #1 for Data Science leaders right now is: framing and narrative. Executives don’t have a
In 1874, Eads Bridge became the longest bridge in the world and the first to be
AI Products require a “Design for Adoption” strategy. 5 ways we avoid adoption failure 👇 1.
Great leaders see around the corner of today to the challenges of tomorrow. They talk to
The ability to scale your startupis driven more by… How you build the customer base,than how
“Well guys, we live to fight another day.” This how a former CTO finished each workday.
AI Product Team: “We have a working ML model! 🎉” Savvy execs know… There are 3
AI is an expensive, high-performance material — much like Carbon Fiber in the physical world. Carbon
Early-stage product leadership requires you to influence (without authority). One of my favorite leaders had a
There’s nothing wrong with “old feature innovation” if that’s what drives impact. Customers need to actually
The best question executives can ask their AI teams right now: “Are users finding our AI
The difference between users and buyers of a product. In 2009 I launched a medical device
It is really hard to improve upon Ken Norton’s list. Ken Norton: Books for Product Managers
I also come from a non-technical background. I spent my early career in Management Consulting and
Oh boy. I remember those days. I had just landed my first “real” product job. I’d
Product Management is especially prone to odd ball tasks that don’t fit neatly in a job
There are three things you want to optimize for in your first PM job: exposure, mentorship
Great question — I hope others jump in on the conversations. Mistakes are such a rich
© Josh Tong. All rights reserved.