Skip to content
01

CASE / Client project

Ecommerce / specialist retail · Specialist online retailer (name withheld)

Order, supplier and customer operations on autopilot

Every order is logged, split by supplier, chased and communicated automatically, with customer emails held back whenever the data to support them is missing.

n8n · Order intake from the online store
The real n8n workflow for order intake from the online store, close up, with each app and step labelled.
The real workflow · anonymized client project
Client
Specialist online retailer (name withheld)
Industry
Ecommerce / specialist retail
Status
Delivered
Stack
n8n, Wix, Google Sheets, Gmail

01 / The problem

  1. 01Orders arrived from the online store and were tracked by hand in a spreadsheet.
  2. 02Several makers fulfil orders, each with its own order numbering and email expectations.
  3. 03Customer updates depended on someone remembering to send them, with the right tracking number.

02 / Goals

  • One master record of every order and every supplier line.
  • Customer and supplier emails drafted automatically, never with invented details.
  • Nothing depends on the freelancer after handover.

03 / The solution

A set of n8n workflows covering intake, reconciliation, customer drafts, supplier follow-ups, a daily digest and central error handling, all feeding one Google Sheets master record.

01

No-fabrication drafting

Tracking numbers are a hard requirement: a draft that needs one is held, not sent. ETAs are soft: the sentence that carries one simply disappears when it is blank.

02

Supplier-native references

Supplier emails use each supplier's own order reference and never re-list orders already sent, because suppliers asked for exactly that.

03

Safe migration

A stale-draft threshold stops historical orders from triggering customer emails when old records are filled in.

04

Tested against real data

Workflow logic is exercised by automated tests that run the production code against real, sheet-shaped fixtures.

04 / The build

The real workflows, step for step.

These are the production n8n workflows behind this project, redrawn from its real structure. Client names, data, credentials and settings are removed; every step and connection is real.

Every hourScheduleConfigSet fieldsWix: Search OrdersWixRead ExistingOrdersGoogle SheetsIf NewIfAppend Order RowGoogle SheetsLog: IntakeGoogle SheetsManual: ForceResyncManual runRead SettingsGoogle SheetsPlan FetchCodeIf RunIfNormalize OrdersCodePrepare LogCodeBuild SummaryCodeLog: Run SummaryGoogle Sheetstruefalsetrue
n8n15 steps
ANONYMIZED CLIENT PROJECT

Polls the store every hour, skips orders already on file, normalises new ones into the master order sheet and logs a run summary. A manual trigger forces a full resync.

Rendered from the production workflow export. Labels anonymized; no data, credentials or settings included.

Read all 15 steps
  1. Manual run · Manual: Force Resync
  2. Schedule · Every hour
  3. Set fields · Config
  4. Google Sheets · Read Settings
  5. Code · Plan Fetch
  6. If · If Run
  7. Google Sheets · Read Existing Orders
  8. Wix · Wix: Search Orders
  9. Code · Normalize Orders
  10. If · If New
  11. Google Sheets · Append Order Row
  12. Code · Build Summary
  13. Code · Prepare Log
  14. Google Sheets · Log: Run Summary
  15. Google Sheets · Log: Intake
  • Schedule
  • Wix
  • Google Sheets
  • If
  • Manual run
  • Code

05 / Outcome

  • Delivered as 19 n8n workflows covered by 199 automated tests.
  • Live on the client's own n8n workspace since August 2026.
  • Every order and supplier line now lives in one master record instead of a hand-kept spreadsheet, and every customer email is drafted from it.

06 / Walkthrough

Silent · 10 s
ANONYMIZED CLIENT PROJECTAI customer inbox reader: the real workflow, anonymized. A slow pan across the real customer inbox workflow: new mail is fetched, read by OpenAI and written back to the order sheet.

A narrated walkthrough can be recorded on request; client data stays private.

→ / Next step

Want a system like this one?

Next case studyAI sales agent for marketplace event enquiries