Connect CJ Dropshipping: pull products, publish, then dropship orders

Fill your CJ API Key as api_key in the generic editor, test on the Sourcing side, set pull keywords, pull products into the pool, publish them, and push to Shopify.

The CJ Dropshipping connector is a sourcing connector, not a store connector: there is no store authorization and no shop domain to fill in. You save a CJ credential, pull CJ products into the sourcing pool, review and publish them as your own products, push published products to Shopify, and later place dropship orders back to CJ when sales orders come in.

Prerequisites#

• A CJ Dropshipping account, and a CJ API Key that you can obtain on the CJ platform side.
• Access to Integration → Connections and to the Sourcing pages, where you test the connection, pull products and review the pool.
• A rough idea of what you want to source: the pull keywords you will search with, and — if you plan to list products — a Shopify connection, because Shopify is currently the channel with live listing from the pool.
• For dropship orders later: be ready to choose a default logistics option for this connection. Without a default logistics option, the system will not submit dropship orders automatically.

Get your credentials#

1. On the CJ platform side, obtain your CJ API Key. That API Key is the credential this connection uses.
2. Copy the API Key exactly as CJ gives it to you and keep it somewhere safe. You will paste this value into the connection in the next step.
3. Do not treat your CJ login email as the credential. The email alone is not accepted as a credential — a connection that only has an email will report that credentials are missing.

Fill it into Integration#

1. Open Integration → Connections and create or open the CJ Dropshipping connection. CJ has no dedicated wizard fields, so the connection opens in the generic editor — that is expected, not an error.
2. In the generic editor, add an entry with the key name exactly api_key, and paste your CJ API Key as its value. The key name matters: the system looks the credential up by this name.
3. Alternative: if you already have a CJ access token, you can fill that in directly under the key name access_token instead. The recommended path is still api_key above — one of the two must be present.
4. Save the connection. Do not save only your CJ login email — an email on its own is not a credential and the test will report missing credentials.

Test the connection#

Test the connection from the Sourcing side — that test is the real check. With api_key saved, the test uses that key to exchange an access token with CJ: if the exchange succeeds, the connection is really live. If no real credential is saved, the test instead shows simulated mode and works through a sample catalog. A simulated result is not a real connection — it only proves the flow runs end to end. Learn to tell the two apart before you pull anything: real success names the token exchange, simulated mode says so in the result.

Mapping / pick a template#

For CJ, this step is not a field template — it is the pull scope you set in the connection’s edit dialog. Set your pull keywords there: up to 10 keywords, and use the keyword assistant first to probe how many products each keyword would hit before you pull. You can also add a category on top of the keywords to narrow the pull further. When you pull, the matching CJ products land in the sourcing pool. Review them there one by one and use Publish to create each one as your own product. A published product can then be pushed to Shopify to list it — Shopify is currently the only channel with live listing from the pool, so do not expect the same push on other channels yet.

Activate#

Enable the connection, then go to the Sourcing page and pull products with the keywords you set — enabling alone does not pull anything yet. Pulled products wait in the pool until you review and publish them. Once sales orders come in, you place dropship orders to CJ from the order side: choose the default logistics for this connection first, because without a default logistics option the system will not submit a dropship order automatically. Treat the stock numbers you see for CJ products as indicative only — use them to judge whether a product is worth sourcing, not as an exact count you can promise a buyer.

FAQ#

Test reports missing credentials after I filled in my email

Your CJ login email is not a credential. It identifies your CJ account when you sign in, but the connection cannot use it to call CJ — so a connection that only holds an email is treated as having no credential at all, and the test reports missing credentials.

Open the connection in the generic editor, add an entry with the key name api_key and your CJ API Key as the value, save, and test again from the Sourcing side. If you have a CJ access token instead, fill it under the key name access_token — one of the two must be present.

The test passes, but did it really connect?

Not necessarily. When no real credential is saved, the test runs in simulated mode against a sample catalog and still comes back OK. That OK only means the flow works — it does not mean CJ was ever reached, so it is easy to mistake simulated mode for a real connection.

Save a real api_key first, then test from the Sourcing side and read the result wording: a real pass is the one where the system exchanges your api_key for an access token with CJ. If the result says simulated mode or sample catalog, you are not connected yet — fix the credential and test again.

Pulling returns no products, or far too few

The connection can be fine — the pull keywords are simply too narrow, so CJ has few or no products matching them. A pull only brings back what the keywords (and any category you added) actually hit.

Use the keyword assistant in the connection’s edit dialog to probe the hit count for each keyword before pulling. Replace or broaden keywords that hit almost nothing — you can keep up to 10 — and consider whether an added category is narrowing the pull too much, then pull again from the Sourcing page.

A dropship order will not submit

The usual cause is that no default logistics option has been chosen for this connection. A dropship order needs a logistics choice to go to CJ, and without a default the system will not submit the order automatically — it waits instead of guessing.

Set the default logistics for the CJ connection first, then place or submit the dropship order again. Keep the default in place for later orders, so each new sales order can go out without stopping on the same step.