Vendor Reputation Portability in 2026: When Ratings Survive a Market Migration and When They Reset

Vendor reputation does not travel. That is the short version. A vendor who grinds out four hundred escrow sales on one market and then follows the traffic to a successor platform in 2026 lands with a clean slate, a fresh account, and a bond demand from the new operators. The rating history, the feedback score, the accumulated trust, none of it crosses the border with them. This is one of the least examined structural problems in the ecosystem, and it gets worse every time a major platform dies.

Worth stating plainly up front: everything below is for researchers studying marketplace economics and trust architecture. There are no links here and no access instructions. If you are looking for a vendor to buy from, you are reading the wrong article.

Why the reputation ledger is platform-owned

Reputation on a darknet market is not a portable credential. It is a database row on a platform the vendor does not control. The market records every sale, every dispute, every auto-finalize event, and every buyer rating, and it holds the private keys to the wallets those transactions settle into. That is the entire architecture. The vendor contributes labor and risk; the platform owns the record and the money.

Escrow is the mechanism that makes this possible. According to monitoring from SOS Intel, roughly 92% of major marketplaces now run some form of escrow, and the standard flow is boringly linear: buyer deposits, vendor ships, buyer confirms, market releases funds. Multi-signature setups mean no single party can move the escrow unilaterally. But the market always holds its own key, which means the market always has a seat at the table. As one forum post put it after the Abacus collapse, escrow protects you from a vendor, not from the market itself. The operators always hold the keys.

So when a platform dies, the reputation attached to it dies too. The vendor’s transaction history is not a file they can export. It is a row in a database that just went dark, and possibly a set of keys that just walked out the door with the operators.

Migration case study: the Abacus collapse

The clearest recent example is Abacus. The operators took funds and vanished in mid-2025, and the displaced traffic moved quickly. Torzon, which had spent the previous months building uptime and recruiting vendors, absorbed most of the vacuum and is now the ecosystem leader in 2026. That is partly a story about Torzon’s own execution and partly just what happens to whoever is standing when the giant falls.

Here is the part that matters for reputation. Every Abacus vendor who migrated to Torzon started over. Their bond was gone (the context suggests bonds run in the $200 to $500 crypto range, and many are non-refundable). Their feedback score was gone. Their history of reliable escrow transactions, which is what earns the right to request FE (finalize early) status, was gone. A vendor who had proven reliability across hundreds of sales arrived at Torzon as an unknown account with no track record, which is functionally identical to a brand new scammer with a clean registration.

That is a brutal reset, and it has second-order effects. New vendors typically face FE restrictions until they build history through successful escrow transactions. That means migrated vendors spend their first weeks or months on the new platform working under the same constraints as someone who just registered five minutes ago. The only real differentiator they carry is whatever off-platform reputation they can prove, which usually means a PGP-signed message pointing at a now-dead market they cannot authenticate.

What actually survives migration

Not much, and what does survive is fragile.

• PGP keys. A vendor’s signing key is the one genuinely portable asset. If a buyer saved the key from a prior market and the vendor signs the same key on the new one, that is a weak identity link. It is easy to forge trust around a leaked or duplicated key, so it only counts as real evidence if the buyer verified the key out of band previously.
• Off-platform reputation threads. Forum vouches and review aggregators exist, and some vendors lean on them. The problem is that forums are not escrow. A vouch thread has no funds at stake, no dispute mechanism, and no auto-finalize. It is pure assertion.
• Cross-market mirroring. A small number of vendors run shops on multiple platforms simultaneously, which builds parallel histories. That is expensive in bond fees and time and only works for vendors who planned for a migration before it happened.

Everything else resets. Feedback scores, escrow history, FE eligibility, dispute ratios, all of it lives on the dead platform.

The dispute problem has no portable solution

Disputes are where reputation should matter most, and they are exactly where portability fails hardest. A dispute is a formal process where a neutral market moderator arbitrates a disagreement between buyer and vendor, and it can end in a full refund, a partial refund, or payment to the vendor depending on evidence. That entire apparatus is platform-specific.

A new market has no visibility into how a migrating vendor handled disputes on their old platform. It has no data on whether that vendor honored auto-finalize terms, which typically run 7 to 14 days, or whether they pressured buyers into finalizing early. The market cannot import a reputation it never recorded. So it treats every migrating vendor as unproven, which is the only defensible position given the information available.

The vendors who benefit most from this reset are not the honest ones. A selective scammer, someone who scams certain users while maintaining legitimate operations to preserve reputation, gets a free reputation wash every time they migrate. They carry their real track record into a new market as a blank slate, and the platform has no way to know. That is the structural failure at the heart of reputation portability. The honest vendor and the selective scammer look identical on day one.

Flash migration and the automated churn

There is a related dynamic worth flagging. When one hosting provider faces pressure, customers migrate to another within hours using automated tools that sync site content across multiple bulletproof hosts. Criminal infrastructure has gotten good at rapid relocation under pressure. Marketplace reputation systems have not gotten good at anything. The infrastructure moves fast; the trust does not move at all.

Krebs on Security has documented bulletproof hosting providers operating with impunity across jurisdictions, keeping customer infrastructure alive even as law enforcement coordinates takedowns. That means the pipes a market runs on are increasingly portable while the reputation layer stays fixed and platform-bound. The mismatch is only going to widen.

The honest verdict

Reputation portability is close to nonexistent, and I do not see that changing soon. There is no neutral reputation registry that markets trust, no cryptographic primitive that lets a vendor prove a transaction history without a platform vouching for it, and no incentive for platforms to accept imported reputation. A platform that honors a competitor’s feedback scores is just accepting unverified claims from strangers, which is exactly the posture that got users scammed in the first place.

For buyers, the practical read is blunt. After any major market death, treat every vendor on the successor platform as unproven, no matter how loudly they claim a prior history. The Abacus-to-Torzon migration is the template, and it will repeat. For researchers, the more interesting question is whether a vendor’s reputation can ever become a bearer asset, something the vendor holds rather than the platform. Right now the answer is no, and the operators hold the keys.