How an issue gets made.

Overyield is written under a pen name, so the byline can’t buy your trust. The process and the public record have to.

One issue ships Monday with one play: a concrete way to make money with AI, plus a free kit to run it. Here’s the standard every issue has to clear, and what Overyield still doesn’t have.

How an issue gets made

Every issue starts with the tape. Each week we work through 21 sources end to end: 12 podcasts and 9 newsletters across AI, business, and money.

Full episodes, never summaries, because the best material is usually an aside buried in an hour about something else. The AI shows tell you what changed.

The operator and small-business shows tell you who pays for it.

Then the week gets distilled: what actually moved, who reported it, and which numbers survive a second look. The crowded take gets mapped first, on purpose. Consensus is what not to ship. If every AI newsletter will run the same angle by Tuesday, it was never an edge on Monday.

The play comes from working the second-order effects of the week’s biggest real shift. Who quietly gained pricing power.

Who suddenly needs a service that did not exist last month. What a beginner could actually charge for it.

Several candidates get built out, and the one a reader can run inside a week wins.

Then the draft goes through verification. Every number, quote, and causal claim gets checked against a retrievable public source, in a separate pass with no stake in defending the draft.

What fails gets cut, corrected, or relabeled in the open as one operator’s unverified account. The arithmetic gets re-done from scratch, chart by chart.

And every claim the issue makes about Overyield itself, every count and price, gets checked against a canonical record of what this business is today, because the fastest way to lose a skeptical reader is a padded number about yourself.

A second read scores the hook, the argument, the receipt density, the sharpness of the play, and the honesty, and anything that falls short gets rewritten. No new fact may enter after verification.

Last, the gates: every number must match across every file, and the play’s buyer and mechanism must repeat nothing from the last 12 issues. If the week is thin or a check fails, nothing ships. A skipped week with a reason beats a bad issue.

What a play must clear before it ships

• Named sources with dates. A show, a publication, a date. “Analysts say” is not a source.
• Conservative math, rounded down. The modest case, never the best case. Two recovered leads a month, not thirty.
• Who pays whom. The buyer, the pain, the offer, and a price you can charge.
• A stated kill rule. From a recent issue, verbatim: 20 qualified contacts, fewer than 3 conversations, 0 paid pilots, stop.
• An honest part. What’s hard, crowded, or likely to break, written before the play ships rather than after it fails. Plus a bear case: the strongest argument that the thesis is wrong.
• No invented reader results. If nobody has run it yet, the issue says nobody has run it yet.
• A kit that’s a product, not a tip sheet. A worked example, the artifact structure, every template, a day-by-day install plan. If a motivated beginner couldn’t reach a paying result from the free kit alone, it isn’t finished.

The scorecard rules

The board is public at overyield.co/scorecard.

Every call is logged the week it’s made, before anyone knows how it turns out. Nothing is backfilled. Each row carries the original wording, the issue date, and a specific “what to watch” naming the signal that will decide it.

Grading is Hit, Partial, Miss, or Too early, quoting the original wording, never softened into something easier to defend. Nothing is deleted, including a Miss. New evidence goes underneath as a dated update note; the original text stays as written.

22 calls logged. 0 graded. Every row reads Pending, because most calls carry two-to-four-quarter horizons and Overyield started recently. That’s early, and it’s stated rather than hidden. Come back in a year and check the board against what happened.

What we will never do

No income promises. Money is always framed as a price you can charge, never as money you will make.

No invented reader wins. There are none yet, and the scorecard says so in writing.

No paid placement deciding the play. Sponsors and affiliates can fund a free issue, but nobody buys the week’s pick. Affiliate links are disclosed above the fold with #ad.

No fake scarcity. Every published cap and deadline is real and numbered: the founding membership closes at 25 members, not at “limited spots.”

No borrowed authority. No claimed insider access, no invented biography, no imaginary team. Production is AI-assisted, and we say so instead of hiding it. Educational, not financial, legal, or business advice.

Why a pen name

Most credentials in this category can’t be checked. A resume is a claim about the past. A logo wall is a claim about someone else’s work. A scorecard is different: dated, specific, and able to be wrong in public.

Writing anonymously removes the shortcut on purpose. No founder halo, no exits to name-drop. The receipts and the board are the credential, which means the record can’t be curated. If a call ages badly it stays up, because a track record you’re allowed to edit isn’t a track record.

The author writes as Javier. The brand address is javier@overyield.co, and a person reads it.

The honest part

Here’s what Overyield does not have.

Zero graded calls. Every row on the board is still open.

Zero reader results. Nobody has written in with a signed client and a real number, and no invented one will fill the gap.

Zero paying members, so zero member outcomes to report. The community opens when the founding cohort is in, so there’s no active membership to point at yet.

A short archive. The early issues went to the web before the live email cadence resumed, so there’s no send streak to claim.

What does exist: 45 plays in the member Vault, 22 dated calls on the public board, and the process above running every week whether or not anyone is watching.

Read the standard, watch the board, and decide in a few months whether the calls held up.