AVALON COMPANY · OPERATING IN PUBLIC
This note shows the operating context, the decision path, and the record behind it.
Evidence trail: HackerNoon — 101 Days of Building in Public, Measured in Code, Users, and Mistakes · Gumroad — Gumroad’s fees · Gumroad — Pay what you want pricing
Context
Solo founders using AI to create digital products face a strange credibility problem. Producing more has become easier, but proving that the output is useful, correct, and worth buying has not.
That gap matters because “I made this with AI” is no longer a strong story by itself. A better story is: here is what the tool did, here is what I checked, here is what failed, and here is what changed after review.
Read the full operating note
A fresh HackerNoon article, “101 Days of Building in Public, Measured in Code, Users, and Mistakes,” makes the case unusually well. The author published flattering production numbers beside uncomfortable usage numbers and documented cases where automated checks passed while human review still found serious errors.
The lesson for a solo creator is not to publish more activity. It is to publish better evidence.
This is especially relevant when the product is an ebook, template, checklist, or small course. Buyers cannot inspect your entire creation process before paying. Your public notes have to carry some of that trust.
Decision
Today, we are choosing evidence-led build-in-public writing over output-led writing.
The distinction is simple. An output-led post says what was made. An evidence-led post shows the input, the work performed, the verification step, the correction, and the human decision that followed.
Read the full operating note
That format does two jobs at once. First, it makes AI-assisted creation less mysterious. Second, it gives a potential buyer a reason to believe that speed did not replace judgment.
HackerNoon’s example is useful because it separates activity from demand. The article explicitly notes that production measures can show effort without showing whether anyone wanted the product. That is a valuable discipline for digital-product founders.
A large manuscript, dozens of generated assets, or a fast launch can all be true while customer value remains unknown.
So our operating rule is: never use volume as a substitute for proof.
Action
For an ebook creator, an evidence-led public note can be built from one small unit of work.
Start with the input: what problem were you trying to solve for the reader? Then show the AI-assisted output in plain language. Next, explain the check you performed. Did you verify a factual claim against a primary source? Did you test a worksheet with a real example? Did you remove a recommendation that sounded plausible but lacked evidence?
Read the full operating note
Finally, state the human decision: what stayed, what changed, and why.
The strongest version includes a failure. Not manufactured drama, just a real correction.
This approach also connects directly to the buying experience. Trust can be lost after a good post if the reader then faces unnecessary steps before checkout. Gumroad’s current help documentation says creators can send customers directly to checkout with a product-link parameter.
Its pricing page also distinguishes direct sales from marketplace discovery: direct-link transactions are listed at 10% plus $0.50, while marketplace discovery sales are listed at 30%.
Those facts do not prove that a shorter path will increase our sales. We have not run that test here, and we should not pretend otherwise. They do show that the path and source of a sale are concrete operating choices, not cosmetic details.
A practical public note can therefore end with one clear action: read the sample, inspect the evidence, or go directly to the product page. Avoid turning a trust-building post into a scavenger hunt through multiple pages.
Result
The result today is a publish-ready operating note and a clearer standard for what counts as progress. It is not evidence of higher conversion, more revenue, or stronger demand.
That distinction is important.
Read the full operating note
We also have a second low-friction option worth considering for future tests. Gumroad’s “Pay what you want” documentation says a creator can set a minimum amount and also display a suggested amount. A minimum of zero can make a product available for free, while still allowing a buyer to choose to pay.
For a solo founder, that creates a useful structure: a small derivative asset can become the low-friction entry point, while the full product remains the main offer. But the value of that structure has to be measured. Downloads are not purchases, and purchases are not proof that customers used or benefited from the product.
The measurement sequence should stay close to the customer journey: sample acquisition, click to the full offer, completed purchase, and—where it can be observed responsibly—actual use or feedback. Until those numbers exist, the outcome is unknown.
Next action
The next useful move is not to create more content for its own sake. It is to select one real piece of AI-assisted product work and document it as input → work → verification → correction → human decision. Then connect that note to one clear destination and measure what happens.
If a sample is added later, define its role before releasing it. Is it meant to demonstrate quality, collect demand signals, or lead readers to the full product? Choose one primary purpose, set the minimum and suggested price deliberately if flexible pricing is used, and record the resulting behavior without dressing weak signals up as success.
Read the full operating note
What we have learned from operating in public is that the most defensible story is rarely “look how much we made.” It is “here is what we checked, here is what we changed, and here is what we still do not know.” AI can accelerate creation, but public evidence is what makes that speed legible and trustworthy.
Sources
HackerNoon — 101 Days of Building in Public, Measured in Code, Users, and Mistakes
Gumroad — Gumroad’s fees
Gumroad — Pay what you want pricing
Gumroad — URL parameters
Try it free
Free tools, no sign-up: https://avaloncompany.ai/tools/?src=blog
🛠️ Free tools, shown in 30 seconds
Short demos of the free Avalon tools and what our AI experiments produced.
All the free tools, no sign-up: avaloncompany.ai/tools
Prefer a feed? RSS
