Lab Escalations: Process Guide

Audience: Agent-facing – Customer Services team members running lab escalations

Purpose: To explain why we chase lab orders, how to run the twice-weekly escalation process end to end, and the lab-specific quirks that catch people out.

Reference: [Lab Handbook – link here]

Why we do lab escalations

Labs sometimes fall behind on production without flagging it to us. If we don't proactively check, a merchant's order can sit well past its expected production time before anyone notices, usually surfacing only when the merchant chases their own customer complaint. Running a regular escalation check means we catch delays early, get ahead of merchant complaints, and keep a clean audit trail (via Zendesk) of when we chased a lab and what they said.

It also surfaces problems that aren't really about lateness at all: address issues, stock/SKU problems, submission failures, and quality holds all tend to show up in the same order lists, so the process doubles as a general health check on outstanding orders.

Cadence

Reviews typically happen twice a week (Tuesday and Thursday). Orders are reviewed lab by lab, not all at once.

The core process, step by step

1. Get the order list. For each lab, you'll have a list of order IDs and placement dates (sometimes broken down by product category, e.g. Framed, Prints, Canvas).
2. Search Zendesk first. Before treating anything as "needs chasing," search Zendesk for each order ID (and OMS ID, if given). If a ticket already exists, read it before acting; it usually explains the delay (address issue, quality hold, reprint, submission error, and so on) far better than a generic status chase would. Never create a duplicate escalation ticket for an order that already has one open.
3. Check the order on the lab dashboard before escalating. Before drafting any chase-up, check the order directly on the lab's own dashboard (see Lab Handbook for login details). If it shows as already shipped and has a tracking number, there's no need to chase the lab at all – instead, update the order on our admin dashboard to complete and add the tracking number. Only proceed to a lab chase-up if the dashboard shows no shipment/tracking yet.
4. Check the SLA properly. SLA is based only on the lab's production time – order placed to lab shipping it out – not delivery/transit time. Look up the lab's processing time in the Lab Handbook and count working days only (Monday–Friday, excluding bank holidays). If an order was placed on a non-working day, the clock starts on the next working day.
5. Watch for the reprint-date pattern. Order lists are usually chronological. If a date appears out of sequence (earlier than the row before it), it's often a reprint that has retained the original order's placement date rather than a genuinely old order. Flag these rather than assuming they're overdue – a quick Zendesk search or Slack search for the order number will usually confirm whether it's a known reprint.
6. Draft the chase-up before sending anything. Always show the draft (recipient, subject, body) for approval before creating a ticket or posting to Slack. If an order is only borderline within SLA but being included for visibility, say so plainly in the email (e.g. "still within SLA as of today, but wanted to flag it now").
7. Create the Zendesk ticket (once approved) with the fields below, then send.

Zendesk ticket requirements for lab escalations

Note: don't add a sign-off (e.g. "Thanks, [name]") to the email body – the Zendesk signature appends automatically, and a manual one duplicates it.

The "DO NOT USE" check

Before sending to any lab contact for the first time in a session, it's worth checking whether that address is flagged. Search the requester via the Zendesk user search API – if the name comes back as "DO NOT USE [lab name]", stop and find the correct current address before proceeding. Sending to a flagged address doesn't bounce or error, so it's easy to miss until a reply never comes.

Lab-specific quirks worth knowing

• IDT, Cloudprinter, and Jondo Global require an L_ prefix on the lab order ID in all correspondence.
• Bay Photo has a specific required format for every email: their own order number (TNS/PWY-prefixed, preferred), our PO number (lab ID), and the customer reference name. This applies to status chases as well as shipping upgrades.
• Redbubble orders can't be searched by order ID in Oneflow – they need to be looked up by the merchant reference instead.
• Venlo, Alton/Farnham, and Charlotte escalations are typically routed and confirmed via dedicated Slack channels (Venlo support, US support) rather than Zendesk email tickets, since these are in-house production sites rather than third-party labs.
• Some labs only report an "average" processing time rather than a fixed SLA (e.g. Spoke, Taylor OnDemand, Circle Graphics, Disrupt Sports) – treat these as a guideline rather than a hard cutoff when deciding whether to chase.

Escalating unusual submission errors

Not every "stuck" order is a lab delay. If an order is failing to submit entirely (rather than just running late), the fix is different:

• XML error (2, 2) usually points to unusual characters, or misplaced data, in the address fields – check address1/address2 ordering and the companyName field before assuming it's a lab-side fault.
• Validation failures / stock or SKU errors are worth raising directly with the lab, including the full order payload, so they can confirm whether the SKU is available or there's a known issue on their end.

A note on how this guide was built

This guide reflects patterns and rules that have built up through day-to-day escalation work, rather than a single fixed source of truth (aside from the Lab Handbook itself, which is authoritative for contacts and processing times). As processes evolve, or as new lab-specific quirks come to light, this article should be reviewed and updated rather than treated as permanently correct.