Why names are data
Google Ads reports, scripts, the API, analytics imports and spreadsheet exports all carry the campaign name and little else about the campaign's purpose. The campaign type, the market, the funnel stage, the product line and the goal are either in the name or they are nowhere a report can reach without a join to a separate sheet somebody has to maintain. A name that carries those fields lets a pivot table group spend by market, a filter isolate brand campaigns, an analytics report split paid traffic by funnel stage, and a new team member read the account without a briefing.
The second reason is consistency across people and time. Accounts outlive the people who built them; an account named by three successive managers to three private systems is an archaeology project. A written convention is a contract that each new campaign is named the same way, and the audit tutorial treats a broken convention as a finding because it predicts other kinds of drift.
The third reason is that names are where mistakes are caught. A campaign called UK_Search_Brand_Exact
that is serving broad matchSettings that decide how closely a search must match a keyword for the ad to enter the auction. is visibly wrong; a campaign called Campaign copy (new) that is serving broad match
is just a campaign. Names that state intent make the account auditable at a glance, which is the
reason this guide treats naming as a control rather than a tidiness exercise.
The fields a name carries
A name is a sequence of fields. The system defines each field, its allowed values and its abbreviations, and the order they appear in. The fields below cover most accounts; a given account picks the subset it needs and keeps the rest out of the name.
| Field | Purpose | Example values | Notes |
|---|---|---|---|
| Market or region | Where the campaign serves | UK, DE, US, APAC, GLOBAL | Use ISO-style country codes; one convention for regions |
| Channel or campaign type | What kind of campaign | SEARCH, PMAX, SHOP, DISP, VIDEO, DEMANDGEN | Fixed list; matches the campaign type setting |
| Funnel stage or intent | What the campaign is for | BRAND, GENERIC, COMPETITOR, RMKT, PROSPECT | The stage of the marketing funnel the campaign addresses |
| Product, service or category | What is being advertised | SOFAS, ACCOUNTANCY, ENTERPRISE, ALL | Mirrors the product taxonomy used elsewhere in the business |
| Audience or targeting qualifier | Who, where it matters | NEWCUST, EXISTING, LOOKALIKE, INMARKET | Only when the campaign is defined by an audience |
| Match type or structure qualifier | How keywords are matched | EXACT, PHRASE, BROAD, MIXED | Search campaigns only; dropped where structure makes it obvious |
| Goal or bid strategy hint | What it optimises | LEADS, SALES, CALLS, TRAFFIC | Useful where several goals share a product |
| Language | Ad language where markets are multilingual | EN, FR, DE | Only in multilingual markets |
| Version or date | Distinguishes relaunches and tests | a launch period code, or TEST-A | Use a period code, not a free-text date |
Three properties make the fields usable. Each field has a closed list of values, documented with the
convention, so that "UK" never appears as "GB" in one campaign and "United Kingdom" in another. Each
value is an abbreviation short enough that the whole name fits in a report column. And a field that is not
applicable is either omitted by rule or filled with a fixed placeholder such as NA, never left as an
empty position that shifts every later field.
Separators, order and length rules
The separator is the character that splits a name into fields, and it must be one that never appears inside a value. Underscores and pipes are the common choices; hyphens are worse because product names and regions contain them, and spaces are worst because every value contains them. Pick one separator for fields and, if a field has sub-parts, a different one inside it. A sheet formula or a regular expression should be able to split any name into its fields without ambiguity.
Order runs from the most general to the most specific, so that sorting names alphabetically groups the account the way people think about it. Market first groups by country; type second groups by channel within country; funnel stage third separates brand from generic within channel. A name that puts the product first groups the account by product, which is the right order only for an account where product is the primary dimension of reporting. Decide the order from the question the reports most often answer.
Length is a trade between completeness and readability. Interfaces truncate long names, and a name whose distinguishing field is at the end is useless once truncated. The rules that work: fixed abbreviations for every value, the distinguishing fields early, and a cap on the number of fields rather than on characters. Case is a convention too — all capitals for field values is common because it separates names from prose — and whatever is chosen is applied without exception.
One convention applied across an account
A UK-based furniture retailer with campaigns in the UK and Ireland adopts the order
market, type, funnel stage, product, qualifier, separated by underscores, with capitalised
abbreviations from a documented list. The account's 23 campaigns are renamed in one pass on a Tuesday
morning after the four scripts and two Looker Studio reports that referenced the old names are listed
and updated. The resulting names include UK_SEARCH_BRAND_ALL_EXACT, UK_SEARCH_GENERIC_SOFAS_PHRASE,
UK_PMAX_PROSPECT_HIGHMARGIN, UK_PMAX_PROSPECT_LOWMARGIN, IE_SEARCH_BRAND_ALL_EXACT,
UK_SHOP_RMKT_ALL_NA and UK_VIDEO_PROSPECT_SOFAS_INMARKET. Ad groups follow with the campaign's
product field and a theme: SOFAS_CORNER, SOFAS_LEATHER, SOFAS_SOFABED. A pivot on the second field
now splits the £41,000 monthly spend by channel in one step, and a filter on the third field shows that
brand campaigns account for 18% of spend, a figure that had previously taken a manual tagging exercise
to produce each month.
Campaign names by type
The same field system produces different names for different campaign types, because some fields do not apply. The patterns below show how the fields combine; the abbreviations are examples, and the account's own list governs.
| Campaign type | Pattern | Example | Fields dropped |
|---|---|---|---|
| Search, brand | market_SEARCH_BRAND_product_match | UK_SEARCH_BRAND_ALL_EXACT | Audience, goal (implied) |
| Search, generic | market_SEARCH_GENERIC_product_match | DE_SEARCH_GENERIC_ACCOUNTING_PHRASE | Audience |
| Search, competitor | market_SEARCH_COMP_product_match | UK_SEARCH_COMP_ALL_PHRASE | Audience |
| Performance Max, product feed | market_PMAX_stage_segment | UK_PMAX_PROSPECT_HIGHMARGIN | Match type, language unless multilingual |
| Performance Max, lead generation | market_PMAX_stage_service_goal | US_PMAX_PROSPECT_ENTERPRISE_LEADS | Match type |
| Standard Shopping | market_SHOP_stage_segment_priority | UK_SHOP_PROSPECT_CLEARANCE_HIGH | Match type; priority replaces goal |
| Display, prospecting | market_DISP_PROSPECT_product_audience | UK_DISP_PROSPECT_SOFAS_INMARKET | Match type |
| Display, remarketing | market_DISP_RMKT_product_audience | UK_DISP_RMKT_ALL_CARTABANDON | Match type |
| Video | market_VIDEO_stage_product_audience_format | UK_VIDEO_PROSPECT_SOFAS_INMARKET_SKIP | Match type; format added |
| Demand Gen | market_DEMANDGEN_stage_product_audience | UK_DEMANDGEN_PROSPECT_ALL_LOOKALIKE | Match type |
| Test or experiment | base name plus a test qualifier | UK_SEARCH_GENERIC_SOFAS_PHRASE_TEST-RSA | Keeps the base so sorting groups it with the original |
Two conventions keep the table workable. The campaign type field always uses the same abbreviation as the campaign type in the interface, so a mismatch between the name and the setting is an error anyone can spot. And an experiment or a relaunch keeps the full base name and appends a qualifier rather than inventing a new one, so history and the test sit together in a sorted list. The account structure guide decides which campaigns exist; the patterns here name whatever it decides.
Ad groups, assets, audiences and conversion actions
Campaigns get the most attention, but the same discipline applies down the hierarchy and to the shared objects, and the reasons are the same: reports, scripts and strangers.
Ad groups repeat the fields that vary within the campaign and omit the ones the campaign fixes. In a
search campaign split by theme, the ad group carries the theme and, where match types are split by ad
group, the match type: CORNER_EXACT, CORNER_PHRASE. In a Performance Max campaign, asset groups carry
the theme and the audience signal: SOFAS_LEATHER_NEWCUST. Avoid repeating the campaign's own fields in
the ad group name; the report already shows them side by side.
Ads and assets are named where the interface allows a name (image, video and some asset types) and otherwise identified by a convention in the asset itself, such as a consistent file name for images and a title pattern for videos: product, format, size, version. Responsive search ad headlines are not named, which is one reason labels exist.
Audiences and customer lists carry the source, the definition and the window:
WEB_CARTABANDON_MONTH style patterns, with the window written as a fixed code from the convention's list
rather than free text. Lists built from customer data say so, and the business rule that defines them is
recorded with the list, because the name cannot hold it.
Conversion actions are the most under-named objects in most accounts and the most dangerous to get
wrong, since bid strategies optimise towards them. A conversion action name should state the event, the
source and whether it counts as primary: LEAD_FORM_WEB_PRIMARY, PURCHASE_WEB_PRIMARY,
PHONE_CALL_ADS_SECONDARY, NEWSLETTER_WEB_SECONDARY. Reports on CPACost of each conversion. depend on knowing
which actions are counted, and a name that makes the answer obvious prevents the common error of a
campaign optimising towards a newsletter sign-up.
Shared objects — negative keyword lists, placement exclusionLists of websites, apps, video channels or app categories where an advertiser's ads must not appear. lists, shared budgets, portfolio bid
strategies — carry the scope and the purpose: NEG_ACCOUNT_BRANDSAFETY, NEG_SEARCH_JOBS,
BUDGET_UK_SEARCH_GENERIC, PORTFOLIO_TROAS_UK_PMAX.
Labels and custom columns vs names
Not every fact belongs in a name. The test is whether the fact is a permanent property of the object that reports will filter on. Market, type, stage and product pass; they do not change, and every report groups by them. Facts that change, facts that are temporary, and facts that apply to objects at any level belong in labels.
Labels are the right place for status and workflow: REVIEW-Q, PAUSED-SEASONAL, TEST-ACTIVE,
OWNER-JANE, NEW-LAUNCH. They can be applied to campaigns, ad groups, keywords and ads alike, they can
be added and removed without renaming, and reports can filter by them. A label is a tag; a name is an
identity. Using a name to carry a status (..._PAUSED) means renaming whenever the status changes and a
history of names that no longer mean anything.
Custom columns handle derived facts: a KPIA metric tied directly to an objective. computed from other columns, such as cost per
qualified lead when qualified leads are a separate conversion action, or a budget pacing figure. They
belong to the report layer, not to the object, and they should be named with the same discipline — a
custom column called CPL (qualified) is clearer than one called Custom metric.
| Fact | Where it lives | Why |
|---|---|---|
| Market, channel, funnel stage, product | Name | Permanent; every report filters on it |
| Match type, audience, format | Name (where it defines the object) | Permanent and structural |
| Status, owner, review flag, test flag | Label | Changes over time; applies at any level |
| Launch period, relaunch version | Name (as a fixed code) | Permanent once set; distinguishes history |
| Derived metrics and pacing | Custom column | Computed, not a property of the object |
| Business rules behind an audience or a label | Documentation | Too long for a name; must be written somewhere |
UTM alignment
Google Ads auto-tagging appends a click identifier that analytics uses to join sessions to campaigns, and for Google Analytics that is usually sufficient. Manual UTMURL tags identifying traffic source. parameters are added where other analytics tools, a data warehouse, or a CRM need campaign information in the URL, and where the account's naming convention should be the source of the values.
The rule is that the campaign name in the UTM is the campaign name in Google Ads, inserted by the tracking template with the campaign-name value substitution rather than typed by hand. That removes a whole class of mismatches and means a rename in Google Ads propagates on the next click. The tracking template can also carry the ad group, keyword, match type, device and network from Google's value substitutions, which is why those fields need not be duplicated in the campaign name for analytics purposes.
Where UTM values are typed by hand — landing pages shared with other channels, offline material, email —
the same abbreviations and the same case rules apply, so that utm_campaign values from Google Ads and
from email sort and group together. The campaign brief guide
is where a campaign's name and its UTM values should be set down before anything is built, so that every
channel uses the same string from the start rather than reconciling afterwards.
Governance
A convention is only a convention while someone owns it. Governance is the small set of rules that keep the system intact.
Who names. One person or role approves names, usually whoever approves new campaigns. Anyone may propose; the approver checks against the documented field lists before the campaign goes live. In agencies this is the account lead; in-house it is the channel owner. Proposed names are easiest to approve when they arrive with the rest of the build in a search campaign build sheet, where the approver sees each name beside the settings it describes.
Where the convention lives. A single document that holds the field order, the separator, the closed list of values per field with their abbreviations, the patterns per campaign type, the label taxonomy and the UTM rules. It is linked from the account's notes and from onboarding. The document changes by proposal, with a date and a reason, not by a new campaign quietly using a new value.
How to rename. Renaming in Google Ads keeps the object's history and settings. The risk is outside the account: scripts that match names, saved reports and dashboards, analytics filters, spreadsheet lookups, rules and automated bidding configured by name. Renames are therefore batched, dependencies listed and updated in the same pass, and a mapping of old names to new kept for anyone reading older reports.
Change history. The account's change history records renames with a timestamp but not a reason. Keep a short log — date, object, old name, new name, why — alongside the convention document. The same log holds additions to the value lists. Six months on, it is the only way to know why a campaign that appears to have launched mid-year is actually a renamed veteran. The author's campaign management expertise page describes how this governance fits a wider operating routine, and the account hub places naming beside structure and auditing as the three practices that keep an account legible.
To keep campaign names consistent with tracking parameters across every channel, the UTM and naming conventions template gives you a shared structure to adapt and document.
Common mistakes
- No written convention. A system that lives in one person's head and leaves with them.
- Open value lists. "UK", "GB" and "United Kingdom" in the same account.
- Status in the name.
_PAUSEDand_OLDsuffixes that nobody removes. - Product first when reports group by market. Sorted lists that do not match how people think.
- Separators that appear inside values. Hyphens in names with hyphenated products; no clean split.
- Unnamed conversion actions. Bid strategies optimising towards something nobody can identify.
- Renaming without listing dependencies. Scripts and dashboards that silently stop matching.
- Typed UTM values. Campaign names that differ by one character between the account and analytics.
- Fields that belong in labels. Owners, review flags and seasonal pauses written into names and never cleaned out when they stop being true.
- Convention changed by precedent. A new abbreviation introduced by one campaign, copied by the next, and never added to the documented list.
