Before pricing my first digital products, I did what I always do when I don't know something: I collected data instead of opinions.

I pulled the bestseller rankings for 14 keywords across the Gumroad marketplace — 108 products, live prices, rating counts. Rating count matters because it's the only public proxy for actual sales volume.

Here's what the data showed, and how it shaped the 9 products I launched on Nov 1.

The price anchors

Notion templates cluster in a three-tier ladder: $18 entry, $79 mid, $149 pro. Nobody wins with a single price — the ladder IS the pricing strategy.

Info books / playbooks sell at $20-90. Under $10 reads as "thin content" in this category.

Design assets run $36-99. Higher than most beginners guess.

The counter-intuitive pattern: winners price below median

The most consistent signal in the data: top products in crowded categories price UNDER the category median. Not because their content is worth less — because they're farming reviews first.

Reviews are Gumroad's discovery currency. The algorithm surfaces them, buyers filter by them. A 0-rating $29 product is invisible; a 50-rating $12 product gets discovered and can raise prices later.

One outcome beats a general guide

Titles of bestsellers almost always promise ONE concrete outcome: a 30-day plan, a specific workflow, a named result. The "Ultimate Guide to X" pattern underperforms everywhere in the data.

Two categories are already free

In 2 of my 14 categories, the top-ranked products cost $0. You cannot build a business competing with free. Knowing which categories to avoid is arguably more valuable than knowing where to play.

How I applied this

My launch prices, all below category medians:

  • AI Workstation Playbook (14 chapters, from a real production system): $19
  • Automation OS Notion template: $14
  • AI-Proof Resume Kit (3 ATS-safe templates + tailoring system): $12
  • The full market data report this article is based on: $9
  • A weekly trend-report subscription: $39/yr
  • And a $59 bundle, which I fully expect to sell nothing until ratings exist

Day 3 result: 0 sales. I'm publishing that number anyway, because the honest version of "building in public" includes the zeros.

If you want the full dataset — all 12 viable categories, the price anchors, and the raw 108 records — it's a 4-page PDF here. LAUNCH50 gets you 50% off this week. The free 2-page sample of my weekly report is here, genuinely $0.