Vendor Vetting in 2026: Bond History, Sample Orders, and the Signals That Predict a Reliable Seller

Checked a vendor last month with 312 sales, 4.97 average, and three weeks on the market. Bond paid, profile looked clean. Two weeks later the shop was gone and disputes were piling up. Ratings alone did not catch it. The pattern underneath did.

Research disclaimer: This is for researchers, journalists, and harm-reduction analysis only. It does not provide access instructions or encourage illegal activity. All observations are based on archived listings, forum disputes, and public reports.

The bond tells you how the market filters, not how good the vendor is

A vendor bond is the entry fee to sell. On most markets it runs $200 to $500 in cryptocurrency [2]. That number is not a quality score. It is spam control.

Some markets make it refundable if you close cleanly. Others keep it. The policy matters. A non-refundable bond means an admin has less incentive to care if a vendor with a short history disappears. A refundable bond creates at least a small cost to burning an identity. You can usually find the rule on the market FAQ. Read it before you read the ratings.

A bond also sets a baseline for selective scams. A $300 bond is trivial to recover if a vendor takes two $400 orders and vanishes. That is the exact shape of selective scamming that forum glossaries warn about, where a vendor maintains legitimate operations while scamming certain users, often on large orders or targeting new buyers [2]. If you see a new shop pushing bulk deals with heavy discounts, do the math. The bond is not collateral for your order.

I treat the bond as a timestamp. When was it paid, and what happened after? A bond paid six months ago with steady sales is different from a bond paid nine days ago with 150 sales that all appeared in the last 72 hours. The second pattern is farming.

History is more than age. Look for friction

Longevity without friction is not trust. Real history has bumps. Disputes, refunds, reships, a bad batch that gets acknowledged. Markets that run proper escrow will have a formal dispute process where a neutral moderator arbitrates and can decide on full refund, partial refund, or payment to the vendor [2]. You want to see that a vendor has been through that process and survived it.

Check three things in the profile timeline.

• First sale date versus bond date. If sales start the same day the bond was paid, every review is from the honeymoon period. If there is a gap, that can be more honest. It suggests they waited for organic orders.
• Order count versus review count. Escrow markets push buyers to finalize after delivery [1]. Not everyone leaves a review, but a huge gap between finalized orders and feedback often means auto-finalize is doing the work. Escrow funds are released automatically after 7 to 14 days if the buyer does nothing [2]. High sales with low detailed feedback during that window should make you pause.
• Finalize Early status. FE Enabled is usually granted only after many successful escrow transactions [2]. It is also the highest risk setting for a buyer, since you release funds before you verify anything. A vendor demanding FE on small or first orders has not earned that trust. A vendor who has FE privileges but still offers escrow for new buyers is signaling the right way.

I also look for continuity. Vendors who survive market closures or exit scams often have to rebuild. That is normal. What is not normal is a vendor with no history claiming five years of experience but no archived reviews, no PGP continuity, no mention in independent dispute threads. Treat that as a new vendor, regardless of what the bio says.

Sample orders are the cheapest due diligence you can buy

If the product allows small quantities, the sample is the test. Not the rating. The sample.

Do not sample to check quality alone. Sample to check process. How was the listing written. Did the shipping options match the origin claim. Did the vendor communicate within the market message system or push you to an off-market channel. Direct deals outside escrow avoid market fees, which typically run 2 to 10 percent [2], but they also remove your only leverage. A vendor who pushes you to direct deal before a single escrow completion is optimizing for your risk, not theirs.

Structure the sample with intent.

• Keep it escrow. Traditional escrow still works like this. Buyer deposits to a market wallet, vendor ships, buyer verifies, then confirms delivery and the market releases funds [1][5]. Newer setups use 2-of-3 multi-signature where buyer, vendor, and a reputation-bonded arbitrator must cooperate to move funds [1]. Either way, you want that third party in place. According to monitoring, about 92 percent of major markets now offer some form of escrow [5]. If a vendor refuses it, that is your answer.
• Keep it small and trackable in your notes. Time from order to mark shipped. Time from shipped to arrival. Packaging notes without self-incriminating detail. Did the vendor provide a plausible handling window or just say “soon.”
• Test the dispute path once. Not by faking a dispute. By asking a question that would require a dispute if it went wrong. How does the vendor handle a missing package or a damaged item. The best vendors have a clear reship or partial refund policy written in the listing. Vague promises mean selective refunds later.

A single good sample does not make a vendor reliable. It makes one transaction reliable. I will usually run two small escrow orders spaced a week apart before considering anything larger. The cost of two samples is lower than the cost of one large selective scam.

Signals that actually correlate with reliability

Forget the badge count. These patterns have held up across markets I have watched, including the long tail of smaller shops that make up the thousands of active vendors documented on multi-vendor escrow infrastructure [1].

1. Detailed, boring reviews

Short five star reviews with no detail are cheap to fake. The useful reviews mention shipping time variance, communication, and how a problem was handled. Look for reviews that span weeks, not hours. A burst of 30 reviews in one day is marketing. Reviews spread over two months with occasional three and four star ratings look human. Markets that score reputation publicly and persistently create incentives to compete on quality [1]. That only works if the scores are not gamed.

2. Listing hygiene

Reliable vendors maintain listings. Prices update with supply. Shipping options are specific. Terms are consistent across listings in the same shop. Sloppy operators copy templates across products with mismatched details. That correlates with sloppy operational security elsewhere.

3. PGP and identity stability

A stable PGP key that signs messages over time is a better anchor than a handle alone. Handles can be mimicked. Keys are harder to fake consistently. If a vendor rotates keys, they should explain why and sign the transition with the old key. No explanation is a flag.

4. Public dispute behavior

Search the market dispute forum and independent forums. A vendor who responds calmly with evidence and accepts a moderator decision, even when it costs them, is behaving like someone who plans to stay. A vendor who threatens, doxxes, or goes silent in disputes is telling you how your dispute will go.

5. Fee and price realism

If every other vendor charges a similar base price plus market commission, and one shop is 40 percent cheaper with no explanation, that discount is coming from somewhere. Often it is coming from your future refund. Sustainable pricing that accounts for market fees around 2 to 10 percent [2] tends to be boring. Boring is good.

Red flags that look green at first glance

Some signals feel reassuring but are easy to manufacture. I have learned to invert them.

Perfect proof screenshots. Forum guidance on proof is blunt for a reason. Cropped screenshots that hide retailer names, order numbers, or timestamps are hiding something. Real proof shows full confirmation details. Stock photos or template screenshots reused across listings fail reverse image checks [3]. For vendors, the equivalent is a wall of five star images with no transaction hash, no order context, no verifiable detail. A Bitcoin transaction without a public hash is unverifiable [3]. The same logic applies to vendor deals. If they cannot show verifiable, independent evidence for a claim, treat the claim as marketing.

Vouches from new accounts. If every vouch comes from accounts created in the last week and only posting in one thread, they are likely one person [3]. Cross-verify across independent forums with established members. This is slow. That is why it works. Reliable vendors accumulate vouches from aged accounts that have criticized other vendors too.

<