The pivot
What changed, and what it cost.
On 2026-08-27 RepoRanker stopped being a peer-review platform and became a board you buy a place on. That deleted a currency 254 people were holding. This page is the accounting: what went, the numbers that decided it, what the conversion paid out, and what the founder took out of it. The full per-account ledger is at the bottom, as a CSV.
What changed
Was
A peer-review platform. You submitted a repo, other developers wrote reviews of at least 800 characters, a review sat in a 48-hour dispute window before it counted, and the reviewer earned 10 credits when it released. Credits bought five kinds of boost. Reviews fed the ranking.
Is
A board you buy a place on. Position in the paid block is total dollars spent and nothing else. Entry is $3 of new money, a placement never expires, and every price on it is public. Rows nobody has bought are ordered by GitHub momentum.
Reviews survive. Any signed-in developer can still write one and anyone can read it. They now earn nothing, unlock nothing, and move nothing. The 800-character minimum, the 48-hour dispute window, the release step, the credit award, and reviewer reputation and tiers are all gone with the economy that needed them.
Credits and boosts are gone. Nothing earns credits and nothing spends them. All five boost products are retired: leaderboard, featured, newsletter, review request, and backlink. Every balance that existed was converted into placement bid dollars, once, and that conversion is the rest of this page.
Why
254
registered accounts
319
reviews ever written
16
accounts that wrote them
307
repos ever submitted
The review economy never reached the size the product needed. 319 reviews came from 16 accounts, against 307 submitted repos and 254 registered users. The ranking formula those reviews fed was ordering a board almost nobody had contributed to, and asking a few hundred people to write 800-character reviews to keep it fed was not going to compound.
The traffic said the review product was not the thing working anyway. Over the 30 days to 2026-08-27 the site took 1,071 sessions, 298 of them from Google. Sorted by Google entry page, the category pages (/languages, /topics, /types) pulled 67 entries. Every individual repo page combined pulled 11. Six to one.
Search finds the ranked lists, not the repos on them. A position on a ranked list is therefore the thing with value, which is the thing now being sold. It is also why the board stayed ranked instead of becoming a wall of paid rows: strip the ordering out and the category pages stop ranking, and the traffic being sold goes with them.
Traffic figures: Microsoft Clarity, 2026-07-28 to 2026-08-27. They are small on purpose to publish. A $3 floor against roughly a thousand sessions a month is a cheap slot on a small board, and saying so is the point.
How the conversion worked
25 credits became $1, spendable only as placement bid dollars. Nothing else on this site can be bought with money, so nothing else can be bought with a converted balance. It ran once and cannot be run back. $778 went to 253 accounts.
| Where the credits came from | Credits |
|---|---|
| Signup grants (254 accounts at 25 each) | 6,350 |
| Badge bonuses | 17,250 |
| Reviews written | 3,460 |
| Spent on boosts | -4,740 |
| Held when the conversion ran | 22,320 |
Of those, 2,870 were the founder's and were destroyed rather than converted, which is the next section. The remaining 19,450 became $778.
| Largest converted balances | Credits | Became |
|---|---|---|
| jipika | 425 | $17 |
| Bhasvanth-Dev9380 | 330 | $13.20 |
| boyter | 295 | $11.80 |
| henon | 285 | $11.40 |
| ozeranskii | 250 | $10 |
Why 25 to the dollar
The rate decides whether the board is contestable on day one. At 10 credits to the dollar the same balances release $1,945 and the largest single balance is $42.50, more than fourteen times the $3 entry price. One holder would have started that far above everyone arriving after them, and the ladder would have been locked before it opened. At 25 to the dollar the strongest converted balance is $17 and $18 takes it. The rate is also symmetric with the retired signup grant: 25 credits became exactly $1.
The $3 floor does not apply to a converted balance. $3 is the entry price for new money through Stripe. A bid funded from a converted balance may be any positive amount. That carve-out exists because 166 of the 253 balances are under $3, and a floor that applied to them would have handed two thirds of the userbase a number they could look at and never spend.
Every account, every balance, all 254 rows, sorted so the zeroed founder row is first: conversion-ledger.csv. It is rebuilt from the credit transaction ledger, which was left intact and read-only, so the totals on this page can be checked against it rather than taken on trust.
What the founder took: nothing
thejaycampbell held 2,870 credits, 6.7 times the largest balance anyone else held. At 25 to the dollar that was $114.80 of bid power, and it would have been the largest holding on the board before a single bid was placed.
It was zeroed, not converted. The board's headline claim is that every price on it is public. A founder quietly holding $114.80 of undisclosed bid power makes that claim false from the first minute, so the balance was destroyed instead of migrated. It is the first row of the CSV, at $0.00.
The board was rounded, and that moved people
On 2026-09-03 every placement was rounded up to the next whole dollar, so the board reads in dollars rather than in the fractions the credit conversion produced. That raised 171 of the 196 rows and added $87.80 nobody paid. Whatever the rounding added is not anyone's money: it was not charged and it did not come out of a balance.
It also moved ranks, and that broke a promise. This site says a placement holds until somebody spends more than you. Rounding collapsed each dollar band into a tie, ties break on who was placed first, and 74 rows moved down without anyone outbidding them. The largest single move was 67 places. Nobody paid for that and nobody was asked. If yours moved and you want it undone, reply to the notice and say so.
The uncomfortable half
The top of the board is jipika/WaifuX at $17, across 196 paid placements worth $803 in total. The founder's $3 has been passed and moved down the board rather than off it, which is the mechanic working as advertised. Nothing outbid it, though. Not one of the rows above it was paid for: every one was funded by balance this platform granted rather than sold, so the order above the founder is granted money and not new money.
For completeness, because "the founder took nothing" should not be read as more than it says: the records hold two other completed Stripe checkouts, both dated 2026-04-25, both on the retired boost product, and both the founder's own. No other account has ever paid RepoRanker money. 195 other rows on the board were bought with a converted credit balance, which is not new money: those credits were granted by the conversion, never sold, and no card was involved. The 2026-09-03 round-up then raised every fractional placement total to the next whole dollar, so on 171 rows part of the printed price was granted by the house rather than paid or converted: nobody was charged that difference and no balance was debited for it.
Related reading: Disclosure rules · The placement ladder · Ranking rules · The conversion ledger (CSV)