Most naming conventions are born on a single platform. Someone on the team gets frustrated with messy Meta campaign names, builds a format that works, and rolls it out. It's a good instinct. The problem comes six months later, when the same convention needs to work across Google Ads, DV360, CM360, TikTok, and LinkedIn — and it doesn't quite fit any of them.

The platforms differ in meaningful ways: different entity hierarchies, different character limits, different reserved characters. A convention designed in isolation for one platform will either break on another or require so many exceptions that it stops being a convention at all.

This guide covers the practical design principles for building a naming convention that works universally — while respecting the constraints of each platform you run campaigns on.

The Anatomy of a Convention

Before getting into platform specifics, it's worth defining the components that make up any naming convention. Getting these right at the design stage saves significant rework later.

Dimensions are the categories of information encoded in the name: brand, market, campaign type, audience, creative format, date period, and so on. The number of dimensions you need depends on your analytical requirements — every dimension you include is a dimension you can slice performance data by later. Common setups range from 5 dimensions for simpler taxonomies to 15+ for complex multi-brand, multi-market operations.

Separators are the characters between dimensions. The underscore (_) is the most widely used because it's supported across all major platforms and is visually clear. Hyphens (-) work as an alternative but can cause confusion if any dimension values contain hyphens (e.g., date ranges like Q1-2025). Pipes (|) are occasionally used but poorly supported in some platforms and prone to parsing issues in spreadsheets.

Ordering determines which dimensions appear first. The strongest approach is to lead with the dimensions that define the broadest context — typically brand, market, and platform — and progress toward the most specific: audience, creative type, version. This makes names scannable at a glance, which matters when you're looking at a list of 200 campaigns in an ad platform UI.

Value codes are the standardised short forms for each dimension value. NA for North America, AWR for Awareness, PROS for Prospecting. Codes should be unambiguous (no overlap between dimensions), consistent in length where possible, and short enough to respect character budgets. A common mistake is using full names for some values and abbreviations for others — this inconsistency defeats the purpose.

Prefixes and suffixes are optional fixed elements added to the start or end of names — a year prefix like 2025_, a version suffix like _v1, or a platform identifier. These can be useful but add character count, so use them deliberately.

Platform-by-Platform Constraints

Each advertising platform imposes its own structural constraints on naming. Designing a universal convention means understanding where these constraints overlap and where they diverge.

Meta (Facebook / Instagram)

Meta's entity hierarchy runs Campaign → Ad Set → Ad. Campaign names can be up to 400 characters — generous enough that character limits are rarely an issue. The UI and API both accept underscores, hyphens, and most special characters without issue.

The main design consideration for Meta is the Ad Set level. This is where targeting, placement, and budget decisions live, so naming at this level typically needs to encode audience and placement dimensions that don't appear at campaign level. Dimension inheritance — where brand and market values set at campaign level flow down to ad set and ad names — keeps things consistent without redundant selection.

One practical note: Meta's reporting UI truncates long names in certain views. While 400 characters are technically available, keeping names under 80–100 characters ensures they remain readable in the Ads Manager table view without horizontal scrolling.

Google Ads

Google Ads follows a Campaign → Ad Group → Ad hierarchy, with campaign names accepting up to 255 characters. The platform is permissive with characters — underscores, hyphens, and spaces all work.

The complexity with Google Ads comes from campaign type variation. A Search campaign, a Performance Max campaign, and a Display campaign serve fundamentally different purposes and often need different dimension combinations. Some teams solve this by making campaign type one of the leading dimensions (e.g., Brand_UK_Search_Brand-Terms_Exact), which makes the naming pattern self-documenting when scanning campaign lists.

Ad Group naming is especially important in Search campaigns, where the ad group typically encodes keyword theme or match type information. This means your convention may need a different dimension set at ad group level for Search versus Display — a common source of friction if the convention wasn't designed for multi-type use from the start.

DV360 (Display & Video 360)

DV360's hierarchy runs Insertion Order → Line Item → Creative, and the naming context differs from self-serve platforms. Insertion Orders often map to media plan line items, so the naming convention needs to bridge campaign planning and execution.

Character limits are generous (up to 240 characters for most entities), but the real constraint is structural. DV360 campaigns frequently involve programmatic deal IDs, audience segments, and frequency parameters that need to be identifiable from the name. Many teams include deal or package identifiers as a dimension — something that isn't relevant on Meta or Google Ads.

If you're running DV360 alongside CM360 (which is common in programmatic setups), the naming conventions need to be designed together. CM360 campaigns and DV360 campaigns often reference the same underlying initiative, and the names need to map to each other for trafficking and reporting reconciliation.

CM360 (Campaign Manager 360)

CM360 has the most complex entity model of any major advertising platform, and it demands the most disciplined naming. The hierarchy runs Advertiser → Campaign → Placement → Ad, but that's only part of the picture.

Sites exist at the account level and are shared across campaigns — they're not nested under a single campaign but rather assigned to placements as needed. Creatives exist at the advertiser level and are similarly shared, assigned to ads across multiple campaigns. This non-linear model means that Site and Creative naming can't follow the same parent-child inheritance pattern as Campaign → Placement → Ad.

For naming convention design, this has two implications. First, Site and Creative names need to be self-contained — they can't rely on context from a parent campaign because they'll be used across multiple campaigns. Second, Placement and Ad names should encode enough context to be meaningful on their own, since they'll appear in reports alongside placements and ads from other campaigns.

Character limits in CM360 are tighter for some entity types (Placement names, for example, have practical limits around 256 characters), and the trafficking workflow involves moving between CM360 and verification/brand safety tools where names need to remain intact.

TikTok Ads

TikTok follows the standard Campaign → Ad Group → Ad hierarchy, with campaign names up to 512 characters. The platform is relatively permissive with naming — underscores and hyphens both work, and there are few reserved characters.

The TikTok-specific consideration is ad format diversity. Spark Ads (which boost organic posts), TopView placements, and standard In-Feed ads have different creative requirements and performance dynamics. Many teams encode the ad format or placement type in the ad group or ad name — which requires the convention to accommodate TikTok-specific dimension values alongside universal ones.

For teams running TikTok as part of a broader media mix, the key is ensuring that the TikTok convention uses the same dimension codes as other platforms. If PROS means Prospecting on Meta, it should mean Prospecting on TikTok. If the date format is Q1-2025 on Google, it should be Q1-2025 on TikTok. The consistency of values across platforms is what enables cross-platform reporting.

Designing a Universal Convention

With the platform constraints mapped, the design challenge becomes: how do you build one convention that works everywhere?

Start with the lowest common denominator. The tightest character limits, the most restrictive separator rules, and the shortest practical name length across all your platforms define the envelope your convention needs to fit within. If CM360 has a practical limit around 256 characters and you're using it, that's your ceiling — even though Meta allows 400.

Separate what's universal from what's platform-specific. Some dimensions appear at every entity level on every platform: brand, market, campaign type, date period. These are your universal dimensions — they should use identical codes and identical ordering everywhere. Other dimensions are level-specific (audience at ad set level, creative format at ad level) or platform-specific (deal ID in DV360, match type in Google Search). Design your convention in layers: universal core plus platform-specific extensions.

Map dimensions to entity levels, not just platforms. A common mistake is designing a flat naming convention and applying it uniformly to campaigns, ad groups, and ads. In practice, each entity level serves a different analytical purpose. Campaign-level names encode the strategic context (brand, market, objective, period). Ad group-level names encode the tactical context (audience, targeting, placement). Ad-level names encode the creative context (format, variant, version). Your convention should reflect this hierarchy.

Build in extensibility. Conventions evolve. New platforms get added. New dimensions become necessary (a regulatory market identifier, a new audience taxonomy). The convention should accommodate additions without requiring a retroactive rename of everything that already exists. The simplest way to achieve this is to keep dimensions modular — adding a new dimension appends to the name rather than restructuring it.

Apply the parse test. This is the single most important validation for any naming convention: can you reliably decompose a name back into its individual dimensions? If a name is Acme_UK_Awareness_Prospecting_Meta_Q1-2025, can a parser (whether regex, a script, or an API) unambiguously extract Brand=Acme, Market=UK, Objective=Awareness, Audience=Prospecting, Platform=Meta, Period=Q1-2025? If any dimension value contains the separator character, or if the segment count varies across names, the convention has a design flaw that will cause analytics problems downstream.

Tuxonomy approaches this by managing dimensions and rules in a centralised platform — one dimension library feeds platform-specific rules, each respecting that platform's entity hierarchy and constraints. A single taxonomy generates compliant names for Meta, Google, DV360, CM360, and TikTok simultaneously, with the parse logic built in from the start. The rules engine handles the platform-specific variations so that the team works from one set of dimensions, not five separate spreadsheets.

Your Convention Checklist

Eight questions to evaluate whether your convention is production-ready:

1. Is every dimension value unambiguous? No code should appear in more than one dimension's value list. If NA could mean North America or Not Applicable, you have a collision that will break parsing.

2. Does the separator appear in any dimension value? If your separator is _ and a dimension value is brand_awareness, your parser can't tell where one dimension ends and the next begins. Either change the separator or the value.

3. Is the segment count consistent across all names using the same rule? Variable-length names are extremely difficult to parse reliably. If campaign-level names sometimes have 5 segments and sometimes 7, something is wrong.

4. Do all platforms accept your chosen separator? Test this in the actual UI and API, not just in documentation. Some platforms handle special characters differently between the web interface and the API.

5. Are your names under the character limit of your most restrictive platform? Check the limit for every entity type you name, not just campaigns.

6. Can a new team member generate a compliant name without memorising the convention? If the answer is no, the convention is too complex for manual use and needs a generation tool.

7. Are dimension codes consistent across platforms? If the same audience is RETARGET on Meta and RTG on Google, your cross-platform reporting will require a translation layer that shouldn't be necessary.

8. Can you add a new dimension without renaming existing campaigns? If adding a dimension breaks backward compatibility, the convention structure needs revision.

If your convention passes all eight checks, it's well-designed. If it fails on more than two, the convention is likely generating downstream data quality issues that you may not be aware of yet.