Rep links rot. Posts get deleted, agents change routes, Discords get purged, and the spreadsheet somebody maintained in 2024 becomes a list of dead blue text. So the useful way to compare MuleBuy and SugarGoo is not to guess at their warehouses — we have not seen either — but to ask which of their links is still worth having in a bookmark folder years after you saved it.
The two of them happen to sit at opposite ends of that spectrum, which makes them a genuinely instructive pair rather than an arbitrary one.
Two ways to name a product
MuleBuy's buy link is minimal: a numeric marketplace item id, a capitalised token naming which of Weidian, Taobao or 1688 it came from, a referral parameter. The whole address of a product is two values.
SugarGoo's is maximal and hidden: the entire encoded source URL, plus a member id, living after a hash. A hash fragment is never sent to a server — the browser keeps it and hands it to the JavaScript on the page — so on SugarGoo the part of the link that identifies your item is read by their front-end router rather than their infrastructure. The SugarGoo review sets out that mechanism in full.
They are not quite pure opposites, and the exception is worth knowing. When a source URL will not parse cleanly, MuleBuy's link falls back to carrying the whole encoded URL too. So MuleBuy is id-first with a URL-shaped safety net; SugarGoo is URL-only, all the time.
The durability test
Here is what actually decides whether a saved link is recoverable later, and it is not which agent you saved it from.
- Is the item still identifiable from the text of the URL alone? On MuleBuy, the id is sitting there in plain sight. On SugarGoo, the original marketplace URL is inside the product parameter in encoded form. Both pass — which is the answer people find surprising.
- Did the link survive being copied? MuleBuy's parameters are in an ordinary query string, the part of a URL that everything preserves. SugarGoo's are in a fragment, and fragments are the part that link shorteners, chat previews and redirect wrappers treat as disposable. This is where the two genuinely diverge, and it is a transit problem rather than a storage problem.
- Does the agent still use that route? Neither you nor we control this. Routes change, and when they do, every link built on the old one dies at once regardless of grammar.
Point three is why the honest advice is the same for both agents and slightly boring: save the original Weidian, Taobao or 1688 product URL, not the agent link. The source link is the durable artefact. The agent link is a rendering of it, and renderings expire.
When you only have the agent link
Which is most of the time, because that is what people paste. Both shapes are recoverable, and the link decoder does it for either: it reads the item back out of the URL's own structure with no network call, so a MuleBuy link from a rotted post and a SugarGoo link with a mangled tail both resolve into a working buy link for all eight agents at once. A converter that handed you one agent would be an advertisement; handing you all eight is the only honest way to run a monetised list.
The open question on one side only
MuleBuy's referral parameter sits in an ordinary query string, where a server sees it. There is nothing ambiguous about that placement.
SugarGoo's member id is written inside the hash, on the reasoning that their client router is what reads product parameters there. Nobody here has run the test that would settle whether attribution actually survives that placement, and this site's own test suite deliberately refuses to assert either reading.
We are surfacing that because it is the sort of thing an agent comparison normally hides. It has no effect on your price either way. It is a hole in our knowledge about our own links, and a page that conceals its open questions has not earned trust on the ones it answers.
Cost: identical shape, and no figures from us
Both build an order the way every proxy buyer does — the item plus its domestic Chinese leg to the warehouse, a service or purchase fee, storage that is free for an opening window and then charged per day, and international freight that does not exist as a number until the box is packed and weighed on chargeable weight. Add-ons sit on their own lines at both.
No percentages for either appear on this page. We have not audited them, they change without announcement, and a stale rate stated confidently gets copied across twenty sites for two years after it stopped being true. Their own fee pages bind, and how agent fees work explains why the line these two would advertise against each other is rarely the line that decides your total.
Worth adding, since MuleBuy's link grammar is shared with two other agents in this set: a common front end tells you nothing about warehouses, forwarder contracts or dispute handling. That argument is made properly in CnFans vs MuleBuy.
What is missing, and it is most of it
Nobody writing this has placed an order through either agent. There is no timing data from order to warehouse, no comparison of QC photo counts or quality, no dispute or lost-parcel record, no payment success rates by country, and no matched basket — same items, same weight, same destination, same week, all-in — priced at both. Until that exists, neither is ranked, alongside the other six.
The corpus behind this site is dated Reddit comments about sellers, quoted verbatim and linked back to their threads on each seller page. Those buyers are grading garments, not the proxy who forwarded the box, and building an agent rating out of that material would mean fabricating the absent part. The rule the method page applies to a seller with no votes applies here: not knowing is not a grade.
If you have to pick one this week
Check what each currently accepts as payment from your country, pull a shipping estimate on a realistic weight to your own address at both, and read both fee pages the same afternoon. Then buy two or three cheap things rather than one expensive one, and inspect the QC photos while the item is still in China and still swappable — how to buy reps has the order of operations.
Both agent links this site emits carry a referral code, disclosed in the methodology, which never changes your price and is exactly why no winner is named above. The other pairing on the same hash-versus-query fault line is Superbuy vs SugarGoo.