Gorgias Tags Explained (and How to Use Them Well)

What a tag actually does in Gorgias

Per Gorgias' Tags 101 documentation, a tag exists to do two jobs: sort your tickets and automate your workflow. A tag can tell you what a ticket is about (a wismo or return-request label), who it should be assigned to, or which Rule fired on it — all without opening the conversation.

The important thing to understand is that a tag is not just a sticky note. It is a filter key. Once a ticket carries a tag, that tag can drive a View (a saved, filtered slice of the inbox), feed a Statistic (a count in your reports), and act as a condition or action inside a Rule. So the label you choose on Monday becomes the routing logic, the dashboard number, and the automation trigger for the rest of the quarter. That is why tag hygiene matters more than it looks: every sloppy tag is a leaky filter downstream.

Tag management lives in your Gorgias Settings, and it is a permissioned area — only Admins and Lead agents can create, edit, or delete tags. That restriction is a feature, not a nuisance: it is what keeps twelve agents from inventing twelve spellings of "refund."

The three ways to tag a ticket

Gorgias gives you exactly three ways to get a tag onto a ticket, and knowing when to reach for each is most of the skill.

1. Individually. Open a ticket and use the "+ Add tag" control at the top of the conversation. This is for the one-off — the edge-case ticket that does not fit any pattern.
2. In bulk. Inside a View, tick the checkboxes next to several tickets, click "Add Tags", and choose the tag. This is how you retro-fit a label across a backlog — say, tagging every ticket from a botched shipment carrier-delay in one pass.
3. Automatically, via a Rule. Create a Rule that watches for a trigger and applies a tag with no human involved. This is where the real leverage is, and it deserves its own section.

You remove tags the same two ways: manually, by clicking the "x" next to the tag on a ticket, or automatically, with a Rule (for example, stripping a NEW tag once a customer replies).

Auto-tagging with Rules: the real workhorse

Manual tags do not scale; Rules do. A Rule in Gorgias performs an action on a ticket when a trigger and a set of conditions match, and "add a tag" is one of its most common actions. Instead of hoping agents remember to label every WISMO ticket, you write one Rule that tags them all, forever.

Gorgias' Auto-tag Rule best practices offers ecommerce-native templates worth copying:

• "Cancel order" — trigger on the phrase, tag cancellation, and route to your fulfillment agents before the order ships.
• "Urgent order edit" — catch customers who want to change a recent order and tag it for a fast lane.
• Integration-based — in a multi-brand setup, tag by the source integration so each brand's tickets land in the right View.
• Language — detect the customer's language, tag it, and assign to a matching agent.

One subtlety the docs call out: Rules can either append a tag or reset the tag set. The Add Tag Action appends to whatever is already there; the Reset Tag Action removes all previous tags first. Reach for reset only when you truly want a clean slate, because it will quietly wipe context an earlier Rule added.

How tags power Views, Rules, and Statistics

Tagging is only worth the effort because of what tags feed. Three things consume them.

Views. A View is a saved filter over your inbox, and tags are its most natural filter key. Tag every warranty issue warranty, build a Warranty View filtered on that tag, and your specialists get a clean, self-populating queue instead of digging through the firehose.

Rules. Tags are not only an output of Rules — they are also an input. A tag applied by one Rule can be the condition another Rule keys off, letting you chain logic: tag vip on intake, then have a second Rule prioritize anything carrying vip. Pair this with your macros and a tagged ticket can practically triage and answer itself.

Statistics. Per Gorgias' Ticket Insights Statistics docs, the Tags report shows how many tickets in a period carried a given tag — counting both manually applied and Rule-applied tags together. This is how you answer "how much of our volume is WISMO?" or "are returns trending up after the holiday rush?" All roles can view the Tags report, which makes tags the closest thing Gorgias has to a lightweight, self-serve analytics layer. If tags are messy, so is every number that depends on them.

Tag hygiene: the boring part that decides everything

Because tags are load-bearing, a little discipline pays off disproportionately. A few habits that hold up:

1. Adopt a naming convention. Lowercase, hyphenated, prefixed by domain — order-cancel, order-edit, return-request. Consistency makes tags sortable and searchable.
2. Prefer a shallow, deliberate set over an organic sprawl. Twenty well-chosen tags beat eighty accidental ones.
3. Merge duplicates instead of leaving them. Gorgias' merge flow shows you which old tag is being replaced before you confirm, so consolidating wismo / WISMO / where-is-my-order is safe and quick.
4. Lean on Rules for anything repeatable. Humans forget; Rules do not. Every tag you can automate is a tag you can trust in a report.
5. Prune on a schedule. Once a quarter, delete tags nobody uses and merge the ones that drifted.

The honest limits — and where AI tagging picks up

Gorgias' tagging system is genuinely solid: it is transparent, permissioned, deterministic, and wired into Views, Rules, and Statistics without any extra tooling. For structured, predictable signals — the source channel, a specific keyword, a language — Rule-based auto-tagging is exactly the right instrument, and you should use it.

But be clear-eyed about where keyword Rules break down. A Rule that tags on the phrase "cancel order" will miss "I don't want this anymore," "please stop the shipment," and "can you not send it?" — all the same intent, none of the same words. Real customers do not phrase their problems the way your triggers expect, so keyword tagging is either too narrow (it misses variants) or too broad (it false-positives). Tag hygiene fixes spelling; it does not fix the fact that language is messy.

This is the seam where an AI layer fits. Gorgias' own AI ticket tagging feature closes part of the gap: as documented in Let AI Agent autofill tags and ticket fields, you go to AI Agent → Settings → AI ticket tagging, click "Add Ticket Tags", and describe in plain language when each tag should apply; the AI Agent then reads incoming email and chat, figures out the topic, and tags accordingly. It is a real upgrade over keyword Rules — but note two honest caveats. It requires an active AI Agent subscription (the add-on, not the base help desk), and it only autotags tickets the AI Agent itself handles — a ticket that goes straight to a human is not touched. Gorgias bills its AI Agent per automated resolution, so the economics of "tag everything with AI" and "answer everything with AI" are bundled together.

There is another way to run the same idea. AI agents for customer service exist to do exactly the reasoning a keyword can't — and an agent layer can tag by intent across your whole queue, not just the tickets it answers. Macha is one such layer: it runs on top of the Gorgias you already use through a live native connector — it does not replace Gorgias or its tags. You connect Macha to Gorgias, and its AI agents read the full conversation, classify by what the customer actually means, and can look up order or subscription status through a custom tool before deciding how to label and reply. The commercial contrast is worth stating plainly: Macha's credits are consumed per AI action, not per resolution — a genuinely different model from Gorgias' per-resolution AI billing, and one you can size against your own volume on the pricing page. (More on how the two fit together in the Macha–Gorgias integration overview.)

The clean division of labour: keep Gorgias tags as the source of truth for what a ticket is and where it goes, keep Rules for the structured signals they nail, and layer an AI agent on top for the intent-level tagging and answering that a keyword match will always miss.

FAQ

Where do I create and manage tags in Gorgias? Tag management lives in your Gorgias Settings, and it is permissioned — only Admins and Lead agents can create, edit, or delete tags. From there you can create new tags, merge duplicates, and delete unused ones.

What are the three ways to tag a ticket in Gorgias? Individually (the "+ Add tag" control on a ticket), in bulk (select tickets in a View and click "Add Tags"), or automatically via a Rule that applies a tag when a trigger and conditions match.

How do tags feed reporting? The Tags report in Statistics shows how many tickets in a period carried each tag, counting both manually applied and Rule-applied tags. All roles can view it, which makes tags a lightweight self-serve analytics layer — as long as your tag set is clean.

Does Gorgias have AI auto-tagging? Yes. AI ticket tagging lets the AI Agent read incoming email and chat, identify the topic, and autofill tags you describe in plain language under AI Agent → Settings → AI ticket tagging. It requires an active AI Agent subscription and only tags the tickets the AI Agent itself handles.

Can I add AI tagging without replacing Gorgias? Yes. An AI agent layer like Macha connects to Gorgias through a native connector and runs on top of your existing help desk and its tags — it does not replace them. It classifies tickets by intent across your queue and can draft replies too, while Gorgias stays the system of record for tags, Views, and Rules. Macha bills per AI action, not per resolution.

Ready to tag by intent instead of by keyword — and answer the ticket too? Start a free trial of Macha and connect it to your Gorgias in minutes.