Decoding AI Business with Jimmy

Others post revenue screenshots.
I check the books.

Each episode takes apart one real AI business. Every number is sourced and labelled: verified through a payment account, self-reported by the founder, or estimated from stated assumptions.

Are the numbers real?收入可信吗
How did it grow?怎么起来的
Can you do it?你能做吗
Why this exists
A screenshot shows what someone made. Only the books show whether the business is worth your time.

AI has made starting a business cheaper than ever. With it came a flood of revenue screenshots, "$100K a month" headlines and copy-this-and-get-rich tutorials.

None of them answer the questions that matter. Who reported that number? Where are the costs hiding? Take the founder away, and how much business is left?

I've worked at Fortune 500 companies, run e-commerce operations and started my own ventures — some worked, some didn't. The lesson that stuck: judge a business by its books, not its screenshots.

So here's the job. Each episode picks one real AI business, finds the numbers that can be verified, labels what's verified, what's self-reported and what we estimated, and then answers as an operator would: could you do this, and how would you do it differently?

No "sure things", no investment advice. Just a method for checking the books, and numbers you can recalculate yourself.

— Jimmy

Jimmy
About Jimmy

I'm Jimmy.

I bring 25+ years of multinational management and startup experience across China and Australia, with deep work in e-commerce operations, business strategy, digital marketing and data analysis. Today I focus on AI startup advisory and business planning, helping founders read the numbers clearly and choose the right direction in this AI wave.

That's why I read a business by its books first, then ask whether an ordinary person could actually do it.

  • 25+ yearsMultinational management and startups in China and Australia
  • E-commerce · StrategyFrom strategy through execution, online and offline
  • Digital marketing · AnalyticsJudging businesses by data, not stories
  • University of SydneyMaster of Project Management

More about me →

How we check the books

The same method, every episode. Full method →

Verified

Payment-account data, company filings, official price pages. The hardest evidence.

Self-reported

What founders or companies say. Recorded as-is, clearly marked unverifiable.

Third-party

Media and data platforms. We name who, when and how.

Our estimate

Calculated from public prices and stated assumptions. You can redo the math.

One email per teardown

One real AI business, three questions, every number sourced. One email per episode. No ads. Unsubscribe any time.

By subscribing you agree we keep your email as described in our privacy notice, only to send this newsletter.