Naming convention template for B2B SaaS

The naming convention with three segments a subscription business needs: which slice of the ideal customer profile a campaign reaches, where in the funnel it acts, and which quarter it belongs to.

Everything else is the general convention, unchanged. One row per segment, fixed positions, a closed list of values with an owner, one separator and a change rule that does not break reporting history.

Adopt it before the first quarter's demand generation is built. The grouping this business reports on is segment by stage, and an account named by channel and campaign theme cannot produce that view at all.

Use it again when the segmentation changes, which in this category happens whenever the sales team reorganises.

A platform groups an account by campaign and by objective. Neither is the grouping a subscription business reports on, which is segment by funnel stage: what the mid-market awareness spend cost against what mid-market capture cost, across every channel.

Putting both in the name makes that a filter. Leaving them out makes it a monthly reconstruction, usually done in a spreadsheet by somebody who has to remember which campaign was which. The reconstruction is what the convention replaces.

That is also why this version of the document adds only three segments. Almost everything else a team wants in a name — the offer, the asset, the test — belongs at the ad level or in the creative identifier, and putting it in the campaign name makes the name long enough to be truncated.

The ICP segment position is only useful if the same values exist in the pipeline data, spelled identically.

That sounds obvious and it is broken almost universally: marketing writes midmarket, the sales system says Mid-Market, and the join that would have answered what a segment's pipeline cost has to be done by hand with a mapping table that somebody maintains until they leave.

So the allowed-values list for this position is copied from the sales system rather than invented, and the owner for it sits close enough to that system to know when it changes. When sales reorganises its segments, this list changes with it, and the change rule says what happens to campaigns already running under the old values.

The funnel stage position does a second job beyond reporting: it protects spend that cannot be judged on the same timescale as the rest.

An account where awareness and capture campaigns are indistinguishable in the naming will, at some point, be reviewed on a single cost-per-lead table, and awareness will be at the bottom of it. Named apart, they are reported apart, and the comparison that would have killed the upper-funnel spend does not get made by accident.

Keep the list short: awareness, capture, acceleration, expansion covers most accounts. A campaign that genuinely does two of those is two campaigns, because its results cannot be attributed otherwise.

A quarter token makes period comparison a filter. It also invites a specific mistake: updating the token each quarter on a campaign that keeps running.

That is a rename, and a rename breaks the history in most reporting. The rule that works is that the quarter token records when the campaign was created, not which quarter it is currently spending in, and always-on campaigns get an explicit evergreen value rather than a rolling one.

Where a campaign really is quarterly — a push built for a specific target — a new campaign each quarter is the honest structure, and it produces a cleaner comparison with its predecessor than a renamed object ever will.

The largest benefit in this category is not in the ad account at all. It is that the tokens reach the lead record, so a closed deal can be traced back to a segment, a stage and a quarter without anybody joining on campaign names by hand.

That requires the tracking parameters to be built from the same values, and the form or the capture mechanism to store them. Both are usually somebody else's system, which is why the where-it-appears column matters more here than anywhere else: it is a specification for a system outside marketing, and it needs to be agreed rather than assumed.

Check one live path end to end before a quarter starts: click an ad, fill the form, and look at the record. That ten-minute test finds almost every mismatch.

Campaign names in this category get long, because there are more real distinctions to carry.

Order the positions so the ones that survive truncation are the ones you filter on most: channel, segment, stage early; the quarter and any descriptive tail late. Then check the actual limits of the platforms you run, and write them in the notes, because whoever builds campaigns next quarter will not know them.

Where names are consistently cut, shorten the tokens rather than dropping positions. A four-character segment code with a meaning column behind it is a better record than a readable name that cannot be split.

A convention changes how an account gets built, and in this category that is a benefit rather than a side effect.

Once the campaign name has to carry one segment and one stage, campaigns naturally become single-segment and single-stage, which is the structure that lets budget be argued about sensibly. The alternative — one large campaign per channel with everything inside it — makes the naming positions meaningless and makes the budget question unanswerable at the level anybody cares about.

There is a real trade-off, because splitting campaigns spreads conversion volume thinly and some platforms optimise worse as a result. That is a decision worth making deliberately, with the volume in front of you, rather than by default. Write the decision in the notes so the next person knows it was a choice.

Seven checks. One separator everywhere, never inside a value. The ICP segment values are copied from the sales system, spelled identically, with an owner who will hear when they change. The funnel stage list is short and closed. The quarter token records creation rather than the current period, and always-on campaigns have their own value. The tokens reach the lead record and one live path has been tested end to end. Names are set once, with renames recorded. And the truncation limits of your platforms are written down.