UTM Naming Conventions: A Campaign URL Checklist
Create a consistent campaign dictionary, inspect generated URLs, and test the whole tracking journey.
Start with the reporting question
Before producing links, decide what you need to compare: channels, promotions, individual creative placements or all three. A campaign dictionary should make those distinctions clear without assigning a new value to every link. Write down the owner and purpose of each field so collaborators can choose existing names. A URL builder can create syntactically valid parameters from inconsistent inputs, so the naming decision belongs before the export rather than after a report fills with fragmented labels.
Separate source, medium and campaign
A simple newsletter example uses source newsletter, medium email and campaign autumn-launch. Source identifies the referring publication or service; medium describes the channel; campaign identifies the promotion. Content can distinguish hero-button from footer-link while retaining the same campaign. These example values are a local convention, not universal required names. Reserve term for a relevant paid-keyword distinction instead of using it as a spare field for unrelated details that another collaborator will interpret differently.
| Field | Example | Question answered |
|---|---|---|
| utm_source | newsletter | Which source? |
| utm_medium | Which channel? | |
| utm_campaign | autumn-launch | Which promotion? |
| utm_content | hero-button / footer-link | Which placement? |
Example conventions, rather than universal required names. Keep personal information out of URL parameters.
Choose one capitalization convention
Google documents that campaign values are case sensitive: google and Google can be different values. Decide whether your team will preserve entered names or normalize them. Our bulk builder provides preserve, lowercase and lowercase-with-hyphens naming options. Converting Summer Sale with the hyphen option produces summer-sale. Check how this fits existing reports before changing the convention. A normalization setting makes future rows consistent but does not merge historical data already collected under several spellings.
Prepare a small campaign table
Use CSV headers url, source, medium and campaign, with optional content and term. Start with one landing page and two placements so the review is manageable. Keep the destination URL separate from the campaign labels. Quote any cell containing a comma. The bulk builder retains unrelated query parameters and fragments but replaces standard lowercase UTM parameters, including duplicate copies. Review that behavior before using a link that already contains tracking fields or whose exact query string has another purpose.
Inspect the output as a destination
With destination https://example.com/shop?ref=homepage#offers, source newsletter, medium email and campaign autumn-launch, the generated URL retains ref=homepage and the offers fragment while adding campaign parameters before that fragment. Check the decoded values as well as the raw URL encoding. Use a real staging or published destination for the final click test. Do not send signed URLs through a normalizing builder without confirming whether changing the query serialization invalidates their signature.
Keep personal data out of campaign labels
Use a placement or campaign name instead of a customer’s email address, personal name or account details. Google’s Analytics guidance specifically cautions against personal identifying information in campaign parameters. A link can be copied, shared and appear in logs beyond the original message. Review the whole destination query as well as the UTM fields. Removing private details from only utm_content does not fix another parameter that still contains an identifier you should not transmit.
Test collection separately from construction
Click a reviewed link, verify that it reaches the intended page, and inspect the appropriate analytics test or reporting workflow for your configuration. A valid URL does not establish that the analytics tag is installed, that consent conditions permit collection or that a redirect preserves every parameter. Keep a checked sample with the naming dictionary, then build the larger batch. When a campaign report surprises you, compare the published link and landing journey before assuming the generator or attribution model is wrong.