x-octo home Business judgment on AI products
中文

Real-revenue cases

How We Grew Revenue By 60% YoY

Real demand

WorthPoint sells subscriptions to antique dealers, collectors and insurance adjusters who need to price a piece of old stuff against a 530-million-item price database.

Founder interview, self-reported 搜索营销线上广告媒体合作线下寻宝活动
Primary source
View original

01

Real-demand verdict

Real-demand verdict

Real demand

Valuation is an unavoidable step in collecting and resale, and it used to mean paper catalogs, auction visits or asking an expert. A single search replaces that, which is why people keep paying; subscribers grew 40% year over year in 2020.

How it makes money

Subscribers pay a recurring fee for access to the price guide, the marks-and-symbols database and the digital library; the actual price tiers are not disclosed in the source.

What old behavior it replaces

Paper price catalogs, auction-house visits and asking a dealer to eyeball a value.

02

Where the first customers came from

Not stated in the source.

Acquisition channels 搜索营销线上广告媒体合作线下寻宝活动

03

Tactics you can copy

  1. 01Keep buying ads when competitors pull back, and take advantage of the lower CPMs.
  2. 02Treat database depth as the moat: 36 million new price records were added in 2020 alone.
  3. 03Have the founder use the product as a heavy user so real usage drives new-product design.
Moving it to an AI business

Limited AI content here. What transfers is the combination of a proprietary dataset that compounds plus a subscription model, and buying traffic counter-cyclically. For a vertical valuation or search AI product, own the data layer competitors cannot get before worrying about the model.

04

Evidence and limits

Self-reported numbers are unaudited — treat them as leads, not facts
Founder interview, self-reported

Revenue basis · Founder self-reported, unaudited (see source page)

What evidence is missing

The $500K monthly revenue is founder-reported and unaudited, and its relationship to the 60% YoY growth claim is unclear.
Subscription pricing, paying-user count and churn are all undisclosed, so unit economics cannot be assessed.
The AI capabilities (image recognition, machine learning) appear only as plans, with no shipped results.