Macros silently break. Customers notice before you do. Here's the audit that catches broken macros before CSAT slides.
Macros are the workhorses of a Gorgias support team. The refund macro, the address change macro, the shipping status macro. Agents click them dozens of times a day. The macros work invisibly. Until they don't.
Gorgias macros break for boring reasons. Shopify ships an integration update and the refund action stops firing. A team member edits the macro template and introduces a syntax error in a variable. The macro still appears in the agent's panel, agents still click it, but the underlying action fails silently or sends with broken merge tags.
This post is the audit I run to catch broken macros before customers do.
Three flavors of macro breakage, in order of frequency:
Variable breakage deserves its own section, because two of the three causes are documented Gorgias behaviour rather than anybody's mistake. They are working as designed. The design is just easy to get caught by.
First, missing data does not error. Gorgias's documentation states that if the data isn't available for that ticket, the variable is replaced with a blank space. So a shipping status macro sent on a ticket where the order never linked does not fail. It sends, with a gap where the tracking link should be. The customer reads a reply that looks like an answer and contains nothing.
The usual cause is boring: the customer emailed from a different address than the one they used at checkout, so Gorgias never matched them to an order, and every order variable in the macro comes back empty at once.
Second, rules change who the macro thinks it is. Gorgias notes that when a macro is applied by a rule rather than manually by an agent, all current_user variables default to the account owner. If your escalation or out-of-hours macro signs off with the agent's name, every rule-fired copy goes out signed by whoever owns the account. Nothing in the ticket records that this happened.
Any macro that a rule fires automatically and that contains a current_user variable. Those two facts together produce replies signed by the wrong person, indefinitely, with no error anywhere.
Third, custom back-office variables point at an ID. If you pull data from your own system through an HTTP integration, the variable path contains that integration's numeric ID, taken from the end of its URL in Gorgias. Point a macro at the wrong integration and the field has nothing to return, which lands you back at the blank space from the first failure mode.
The cheap guard against all three is to stop typing variable paths. Gorgias puts a variable dropdown at the bottom of the macro editor, grouped by source. Browsing and inserting from there is the difference between a path that resolves and a path that renders as literal curly brackets in a customer's inbox.
The logic-drift case in full: where your return window really lives in Shopify, and why a macro quoting the old one is the last copy to get fixed.
Run this quarterly. Most teams find 2-4 issues per audit.
Gorgias reports show usage frequency per macro. Pull the last 90 days. Sort by volume.
For each macro: create a test ticket, run the macro end-to-end. Check that:
For each top macro, pull 5 recent tickets where the macro was used. Check what actually happened. Are merge tags right? Did actions complete? Did customer respond positively?
Open the macro template. Read the policy reference. Compare to your current documented policy. Discrepancies = update the macro that day.
Macros with zero usage in 90 days are either broken, hidden, or obsolete. Delete or fix.
Build and maintain a refund macro that doesn't break.
The full refund procedure the macro executes: eligibility, partial refunds, and where the two systems disagree.
Try it on one of your own procedures.
Record a process once, AI writes the structured SOP. 3 free SOPs, no credit card.
Even without a full audit, certain usage patterns indicate drift:
Intents drift the same way macros do: 80 percent accurate at launch, 50 percent within six months. The quarterly retraining procedure.
Quarterly for most brands. Monthly during high-change periods (BFCM, major product launches, helpdesk migrations). The audit takes 2-3 hours for a typical Gorgias account with 30-50 macros.
Run macro audit and Klaviyo flow audit in the same week. Both involve looking at active automations and catching drift. Combined they take half a day.
Pull your Gorgias macro usage report. Identify your top 5. Test each end-to-end this week. Document what you find. Fix the worst issue first.
If you want help building maintainable Gorgias workflows, ReccordSOP records the actual macro setup process and detects when it diverges from documented.
Refund macro, intent training, escalation rules, chatbot configuration.
2-3 hours for an account with 30-50 macros. Half a day if you also audit views, intents, and routing rules.
Refund macros after Shopify integration updates. The refund action fails silently more often than you'd expect.
Archive after 6 months unused. Delete after 12. Reduces clutter and the chance of an agent picking the wrong macro.
Set Slack alerts on Gorgias usage anomalies (sudden 30+ percent drop). Most macro breakage shows up as a usage drop within a few days.
3 free SOPs to start. No credit card required. See if drift detection keeps your docs honest.
Start for freeI built ReccordSOP after watching too many DTC ops teams lose months to undocumented workflows. These SOPs are battle-tested with Shopify operators running $1M to $50M brands.
Last reviewed September 20, 2026
Most brands set up Klaviyo flows once and never revisit them. Here's the audit framework to keep them earning.
Most SOPs are wrong within 90 days of publishing. Here's how to detect it before it costs you a customer.
A refund starts in your helpdesk and ends in your store. The seam between them is where refund procedures quietly break.
We use essential cookies for sign-in and a small amount of analytics to improve the product. Privacy policy.