QAD and Mailchimp integration

What moves between them

The main flow runs from QAD into Mailchimp. ml-connector reads QAD customer accounts and writes them into a Mailchimp audience as list members, with the same records added as e-commerce customers under a store tied to that audience. The QAD item catalog is mirrored as Mailchimp products so order line items reference a real catalog. In the other direction, Mailchimp e-commerce order totals, currency, and line items are read and posted into QAD against the matching customer and sales records for attribution, since orders are the only finance-shaped data Mailchimp holds. Polling cadence is set per object to fit how often each side changes.

How ml-connector handles it

ml-connector stores both credential sets encrypted. On the Mailchimp side it resolves the data center prefix at connect time, from the suffix on the API key or the OAuth metadata endpoint, then sends the key as HTTP Basic or the OAuth token on every call against the dc base URL. On the QAD side it accepts the full tenant URL per customer, since QAD publishes no shared base address, and validates entity paths against that instance. Because QAD cloud is pull-only and Mailchimp exposes no finance objects to subscribe to, both systems are polled on a schedule using offset and count pagination rather than waiting for a push. A Mailchimp store must exist and be linked to an audience before any customer, order, or product write, so ml-connector creates and maps the store first, then upserts e-commerce records by id so a re-run updates rather than duplicates. Members that are cleaned cannot be re-subscribed, so those rows are skipped. Mailchimp caps concurrent connections and returns HTTP 429 when exceeded, so ml-connector limits parallel calls and backs off, and it can also receive unsigned audience webhooks on a secret URL where they help. Every record carries a full audit trail and can be replayed if a downstream call fails.

A real-world example

A mid-sized consumer goods manufacturer runs QAD Adaptive ERP for production, procurement, and finance, and its marketing team runs campaigns in Mailchimp. Before the integration, someone exported a customer list from QAD every quarter and imported it into a Mailchimp audience by hand, so the list went stale between exports and purchase history never made it back to the ERP. With QAD and Mailchimp connected, new QAD customers appear in the audience and as e-commerce customers automatically, the item catalog shows up as Mailchimp products, and order totals flow back into QAD for attribution. The marketing audience stays current and the manual export step is gone.

What you can do

• •Write QAD customer accounts into a Mailchimp audience as list members and e-commerce customers.
• •Mirror the QAD item catalog as Mailchimp products so order line items reference a real catalog.
• •Read Mailchimp e-commerce order totals, currency, and line items back into QAD for attribution.
• •Authenticate Mailchimp with an API key or OAuth token and resolve the data center prefix, and QAD with its tenant-specific session.
• •Poll both sides on a per-object schedule, with store-first setup, upsert dedup, retries, and a full audit trail.

Questions

Which direction does data move between QAD and Mailchimp?

The main flow is QAD into Mailchimp. Customer accounts move from QAD into a Mailchimp audience as list members and e-commerce customers, and the QAD item catalog is mirrored as Mailchimp products. In the other direction, Mailchimp e-commerce order totals and line items are read back into QAD for marketing attribution, since orders are the only finance-shaped data Mailchimp holds.

Mailchimp has no invoices or GL accounts, so what financial data can it share with QAD?

Mailchimp is a marketing platform, not an ERP, so it has no purchase orders, supplier invoices, or GL accounts. Its e-commerce order object is the closest analog, carrying an order total, currency, financial status, and priced line items. ml-connector reads those orders and posts the totals into QAD against the matching customer and sales records, which is enough for revenue attribution but not full AP or ledger postings.

Does the integration rely on Mailchimp webhooks, and how is the data center handled?

Mostly no. Mailchimp Marketing webhooks only fire on audience events such as subscribe and unsubscribe, carry no signature, and cover none of the e-commerce or finance data, so ml-connector reads on a schedule using offset and count pagination, which also matches QAD cloud. The account data center prefix is resolved at connect time from the API key suffix or the OAuth metadata endpoint, then used to build every request URL.