You only notice a naming convention when it is missing: at thousands of ads across markets, mess becomes unusable. Here is how to build a system that scales with you.
A UTM naming convention that scales has three properties: a machine can read it, it encodes only things you will want to filter on later, and everyone applies it without exception. That last part is the hardest and the most important. At a handful of ads, your memory papers over every bit of sloppiness; at thousands of ads across multiple markets, your naming is the only thing keeping your data searchable.
Why does naming break once you scale?
In your home market you can get away with campaign names like summer test v2 final for a long time. You still remember what means what. But every new market, every extra teammate and every hundred extra ads makes that memory worthless. If you want to know how a concept performs across all your markets, you have to be able to filter that from the name. If that information is not structured into it, the question simply no longer exists in your data.
We see this at almost every brand going international. The tracking setup is fine, but the naming grew differently per market, and suddenly a simple question like which hook wins in which country costs an afternoon of manual work. Having produced 15.000+ creatives for brands across 18 countries, we can assure you: this problem never shrinks on its own.
The pain is not limited to analysis either. Your UTMs feed your entire first-party layer: the source of an order in Shopify, the segments in your email tool, the reports you steer on week after week. One sloppy field flows through to every place where you will want to make a decision later.
What do you encode in your naming?
The temptation is to stuff everything into the name. Resist it. Encode only the dimensions you will actually want to filter or group by later:
- Market and language: not the same thing, because you run French in both Belgium and France. Encode both.
- Funnel stage: prospecting or retargeting, so you can separate top-of-funnel performance cleanly.
- Concept: the master concept the ad descends from, so you can compare concepts across markets.
- Variant or hook: which version of the concept, so you can see which opening does the work.
- Format: video, static or carousel, because formats behave differently across placements.
- Launch week: so you never compare young ads with veterans and can trace fatigue back to its start.
Everything beyond that, like inside jokes, file version numbers or the editor's name, does not belong in your UTMs. Every extra freedom is a future inconsistency. If you do want to keep such details somewhere, give them a home in your internal documentation or in the file name of the creative itself, and keep your UTMs reserved for the dimensions your reporting runs on.
What does a scalable convention look like?
The exact shape matters less than the rules behind it. Pick a fixed order of fields, a fixed delimiter, and write everything in lowercase without spaces. A campaign name then reads something like: nl_prospecting_conceptname_video_w12. Boring, and that is exactly the point. Boring means predictable, and predictable means every analysis, dashboard and export can build on it.
Document the convention in a living document with a dictionary of allowed values. What counts as a concept, which abbreviations you use per market, how you name a new hook. New team members and external partners read that document and produce consistent names from day one. Without a dictionary you end up with three spellings of the same concept, and three spellings means three rows in your report that should have been one.
A naming convention is a gift to your future self.
How do you sustain it across thousands of ads?
Discipline is something you organize, not something you hope for. Three things make the difference. First: templates. Nobody types UTMs by hand; names are assembled from fixed building blocks, so mistakes cannot even occur. Second: one owner. One person manages the dictionary and decides on new values. Third: a weekly check. Whoever spends five minutes a week looking at new names catches deviations before they multiply.
Treat your naming as part of your creative workflow, not as admin after the fact. The name is born the moment the concept is created, in the same template where the briefing and the variants live. What starts out right at the source never needs repairing later.
And accept that your convention will need to move house at some point. Brands growing from their first market to creatives in 10 languages discover fields they forgot. That is not failure; build your convention so a field can be added without breaking everything, and migrate at a fixed moment instead of gradually and quietly.
Conclusion
A UTM convention is invisible work that pays you back every week: faster analyses, honest comparisons between markets, and data you dare to steer on. Start small, encode only what you want to filter, and be stricter on discipline than on elegance. Want to see how we structure naming for brands scaling internationally? Book a call and we will gladly look at your setup together.