Building a Campaign Naming Taxonomy That Actually Scales

Why naming conventions break down

Almost every marketing team has a naming convention doc somewhere. A wiki page, a pinned Slack message, a tab in an old spreadsheet. Almost none of them are still being followed a year later. It's rarely that the convention was wrong. It's that one person designed it, once, and it never got built into how campaigns actually launch. New hires never see it. Freelancers and agencies don't have access to it. The moment one campaign gets named inconsistently and nothing happens, the convention stops being a convention.

The anatomy of a good taxonomy string

A campaign naming taxonomy is really just a small, structured schema — a fixed set of fields, in a fixed order, each pulling from a fixed list of allowed values. A typical structure looks something like this:

REGION_CHANNEL_OBJECTIVE_CAMPAIGN-NAME_DATE

The specific fields matter less than the discipline behind them. Every field needs a closed list of valid values. REGION is NA, EMEA, APAC, LATAM, or GLOBAL — never a free-text guess. That's what makes the taxonomy machine-readable later. You can filter, pivot, and roll up performance by region or channel without anyone manually tagging anything after the fact.

Common taxonomy mistakes

• Free-text fields where a closed list should be. "Retargeting" vs. "retargeting" vs. "RTG" mean the same thing to a human and nothing to a query.
• No single owner. When everyone can add a new allowed value on the fly, the list grows without anyone actually deciding it should.
• Too many fields. Past six or seven segments, campaign names get long enough that people start abbreviating inconsistently just to make them fit.
• No enforcement at launch. A convention that isn't checked the moment a campaign goes live only gets enforced after the fact — which usually means it doesn't get enforced at all.

Enforcing it without slowing your team down

The goal isn't adding a bottleneck to campaign launches. It's making the correct name easier to produce than an incorrect one. A generator that builds the string from a set of dropdowns — instead of asking someone to type it from memory — removes most of the drift on its own. What's left is catching the campaigns launched outside that workflow: through an agency, a legacy process, a platform's own native naming.

That part's worth checking automatically, not periodically. In KPI Compass, the Governance page runs every taxonomy string in your uploaded data against your team's schema and flags exactly which field on which string doesn't match. Drift gets caught the next time you upload data — not the next time someone happens to notice. You can run that same check on your own data for free, no login — the Taxonomy Checker takes a CSV and flags exactly which strings break the pattern.