GA4 Custom Channel Group: Separate Ad and Sponsored Traffic

The default channel group in GA4 cannot be edited, but a custom channel group lets you place ad traffic and sponsored content traffic into categories of your own. A custom group is a set of rules on fields such as source, medium and campaign name, and each visit lands in the first channel whose rule it matches. If you run a festival ticketing platform and your banners, sponsored lineup articles and partner links all end up under Referral, the fix has two parts: tag every link consistently, then write those tags into channel rules.

Channel groups and what customizing them changes

According to the Analytics help page on custom channel groups (Opens in a new tab), a channel group is a set of channels, which are rule-based categories of your website's traffic sources. Custom channel groups can be used as dimensions in reports, explorations and audience conditions. So a custom group is not a cosmetic label on one report. It is a classification you can reuse wherever a dimension is accepted.

If you use the interface in another language, menu and field labels may look different. Field names here follow the wording on the Analytics pages.

The raw values stay as they are. Source, medium and campaign name do not change. What changes is the category each combination falls into. That distinction matters because a channel rule can only recognize what the tag says. A link with the wrong medium cannot be rescued by a clever rule, and the better fix is always at the link.

Is a custom group worth building

Practical suggestion: for one campaign, filtering by campaign name is quicker. A channel group earns its place when the same split must appear again and again across reports.

Reading the default channel rules

The default channel group page (Opens in a new tab) lists the rules for manually tagged traffic. It also states that default channel groups cannot be edited in Google Analytics, that channel definitions are not case sensitive, and that Google Ads and Display & Video 360 traffic follow separate rules based on source platform, ad network type and campaign type.

One thing to notice: there is no channel for sponsored content. A medium such as sponsored is not in any of the medium lists in the table, so it does not match those rules. What happens to traffic that matches nothing is explained in GA4 unassigned traffic: causes and fixes.

Why ad and sponsored traffic blur together

A festival ticketing platform might send traffic from three places: a banner on a music news site, a sponsored lineup article on a festival blog, and a partner link on an artist's page. If tagging is inconsistent, these scatter. A banner tagged with medium banner falls into Display. A sponsored article whose link carries medium referral stays under Referral and mixes with organic mentions. A partner link with no tags depends entirely on what the visit's source and medium turn out to be.

This is not a GA4 flaw. It is two business questions meeting one general classification. The default grouping answers a broad question. A platform that pays for editorial placements wants to see paid editorial separately from links people chose to give for free. A custom group exists for that need.

Single channel or separate ones

Both choices can be defended. One combined channel answers the question of total spend on publishers and keeps the report short. Two channels let you compare banner results with sponsored article results. The deciding factor is the person reading the report. If banners and sponsored articles come from different budgets, separate channels make sense. If both come from one line item, a single channel is easier to read.

Splitting a channel later is easier than merging two that were never separated in the tags. So write distinct tags from the start and decide the channels afterwards.

Tags come first, because rules only read links

The Analytics page on collecting campaign data with custom URLs (Opens in a new tab) says that when you add parameters to a URL you should always use utm_source, utm_medium and utm_campaign. It also recommends setting other relevant parameters, especially utm_id and utm_source_platform. Missing UTM parameters result in (not set) values in reporting. Values are case sensitive: utm_source=google is different from utm_source=Google, and the page's best practices point to a consistent lowercase convention.

Lock your vocabulary before writing a single rule: which mediums exist, how sources are spelled, how campaigns are named. Without that, every new link needs a new exception. For tagging in practice, see sponsored post reporting with UTM and GA4 and separating marketplace and social traffic with UTM tags.

Sample tag vocabulary for ticket sales

These values show a format, not a requirement. What matters is that every value is lowercase, spelled one way, and that each link type has a single medium.

Rule order and the first-match principle

The Analytics page states that traffic is included in the first channel whose definition it matches, given the current order of channels in the group, and that you can reorder channels. A visit that fits two rules is assigned only to the one listed first. This is the single key point when writing rules: if a broad rule sits before a specific one, the specific rule never fires.

1. Put the most specific channel first.
2. Prefer medium as the leading condition for each channel.
3. Leave broad catch-all rules for the end.
4. Before saving, take one sample link per channel and trace which rule would claim it, in order.
5. Log every reorder with a date and a reason.

Starter rules to adapt

Practical suggestion: the rows here are a starting point. You choose the condition types in the interface and adjust them to your vocabulary.

The last row works because Default channel group is one of the fields a custom rule can use. Traffic not claimed by your rules can keep its default classification.

Fields, limits and permissions

The page lists the fields a custom group can use: Campaign ID, Campaign name, Default channel group, Manual ad content, Medium, Source and Source platform. A standard property can have 2 custom groups in addition to the predefined group, with up to 50 channels per group. A 360 property, including sub and rollup properties, can have 5. You need Editor access or higher at the property level to create or edit channel groups, and groups with many fields may affect reports negatively. Limits in Google documentation can change, so check the current page before relying on them.

With only two groups on a standard property, spend them deliberately. One could hold the paid publisher split and the other a campaign family split. If you have a spare, use it for testing a draft before touching the production group.

Constraints to plan around

• Custom channel groups cannot be used in the Key events paths report.
• Cost, click and impression reporting is not available for the Manual ad content field.
• Deleting a group is irreversible; affected reports, audiences and expanded data sets show the dimension as unavailable.
• Editing takes effect immediately in reports and segments, but expanded data sets keep the earlier definition.
• Custom channel groups are not available in the BigQuery export schema.

Picking the right field per rule

Medium is the sturdiest starting point for an ad versus sponsored split, because your vocabulary can give each link type exactly one medium. Source depends on publisher names, so the list grows as publishers are added. Campaign name suits a short-lived split but eats the limited channel allowance when treated as permanent.

Making a custom group the primary one

The page says any custom group can be set as the property's primary group in Admin, and that data uses the new criteria from that point forward. Everyone then sees the new classification by default. That is a strong move: colleagues who expect the old channel names may be confused to find different ones.

Practical suggestion: run the custom group alongside the default for a month before promoting it. When comparing against last year, note the change date in the report so nobody concludes that sponsored traffic suddenly appeared.

What applies to past data, audiences and BigQuery

The page states that custom channel groups can be applied to reports retroactively. Today's new group can therefore be used on past date ranges. But if you edit a group that an audience uses, the change is not applied retroactively and only affects user eligibility going forward. Reports look back; audiences look forward.

The BigQuery export schema does not include custom channel groups. If you analyze in BigQuery, rebuild the same logic from source and medium in your query. That is one reason a report and a BigQuery table can show different channel totals.

Channel group or campaign name

Not every split deserves a channel group. If you want one campaign's result, filter a report by campaign name. A channel group fits a split you need again and again, across many reports, audiences and explorations. Ask three questions: is the split temporary or permanent, will it be used in more than one place, and does the channel count stay comfortably under the limit of 50 per group?

When you need many small splits, such as one row per publisher and per campaign, keep the group to broad categories and leave the fine detail to the source and campaign name dimensions. The group stays tidy, and one of your two custom group slots is free for another purpose.

Checks after each edit

The Google page says an edited group applies its new definition immediately to segments, explorations and reports, which can make reports inconsistent with each other. Expanded data sets keep the earlier definition and should not be used after the edit. For audiences, user eligibility changes only going forward.

• Total sessions for one fixed date range are identical before and after the edit.
• The share of Unassigned and Paid Other has not moved without explanation.
• Narrow rules still come before broad ones in the saved order.
• Each audience that uses the group has been reviewed for forward-only eligibility changes.
• Any report built on an expanded data set has been rebuilt.

Practical suggestion: keep a dated change log next to the tag vocabulary. When a number shifts months later, the log answers the first question before anyone opens the interface.

Example scenario: ticketing platform during a festival sale weekend

Example scenario: a festival ticketing platform opens ticket sales for a summer lineup and buys three placements: a banner on a music news site, a sponsored lineup article on a festival blog, and a partner link on an artist's page. After the first week the head of marketing asks what the sponsored article brought. The analyst opens GA4 and finds the article's traffic grouped with unpaid referrals.

The first task is the vocabulary. Did the sponsored article's link carry utm_medium=sponsored? Two of the articles did not. The analyst corrects the missing links. The second task is the group: a Sponsored editorial channel for medium sponsored, placed ahead of broader rules, with a last rule that keeps default labels for everything else. The third task is the test. The analyst follows two sample links, checks that the source and medium arrive as expected in the report, and after processing confirms that the traffic appears in the new channel.

The head of marketing then sees sessions and key events for sponsored editorial on its own line. To read key events properly, see GA4 key event setup. For judging whether an article deserves credit when several touchpoints precede a purchase, attribution models and multi-channel journeys is a useful companion.

Pre-use test checklist

• Is the tagging vocabulary written down, with one medium per link type?
• Are channels ordered from specific to broad?
• Has each channel been tested with a sample link?
• Does a last rule keep default labels for leftovers?
• Do audiences or explorations depend on this group?
• If BigQuery is in use, has the logic been rebuilt in the query?
• Who may edit or delete the group, and who approves a deletion?

Shaping the report around the question

For the question "what did sponsored articles bring", put the custom channel group in the rows and sessions plus key events in the columns, and add the previous period for comparison. Present the result with care. A purchase after a sponsored visit does not mean the article alone caused it, since the visitor may have come through other channels earlier. Add a short note with the group's creation date and the version of the tagging vocabulary.

Decision rule: fix it in the group, the tag or the report

1. If tags are consistent and the default bucket is wrong for your reporting, write a custom group.
2. If the same publisher appears under several spellings, fix the vocabulary first and write the rule second.
3. If utm_medium is absent, no rule can rescue the session, so fix the link.
4. If you only need to isolate one campaign, use the campaign dimension and leave the group alone.

Traffic the default rules cannot see

The default channel group page defines manual-tag rules in specific terms, and the gaps are easier to spot once you read them as conditions rather than as channel names. The Google page also says channel definitions are not case sensitive.

The page describes Unassigned as the value used when no other channel rule matches the event data. It is a signal that nothing matched, not a rule of its own.

Exact match versus regex

The custom channel group page lists conditions such as matches regex and exactly matches (=). Practical suggestion: use exact match when your vocabulary is a closed list, and regex only when one channel must accept several spellings. A regex that is too loose will pull in labels you never meant to include, so test every new pattern against real tag samples before it becomes the primary group.

Debugging traffic in the wrong channel

If a visit shows up where you did not expect it, work through the layers in order instead of guessing. First check the raw values: do the visit's source and medium arrive in the spelling your vocabulary promised? Then check the rule: does the condition match that exact value, or is there a case or spelling difference? Then check the order: could an earlier channel have claimed the visit? Finally check that the report is using the group you think it is, since a different group may be selected.

In every row, the first fix is either a tag or an order. Rules themselves are rarely the problem once tags are consistent and the sequence is right.

Maintaining the tag vocabulary

The real work starts after the group exists. A new publisher, a new link type, or a teammate typing a different medium can break a rule silently. Compare every new link with the vocabulary, and where possible generate links from one shared sheet so nobody types UTMs by hand. Open the channel report every quarter and scan for mediums you do not recognize. For the ticketing platform, a quarterly look would catch a new partner whose link was tagged partners instead of partner, a one-letter difference that would have kept that traffic out of the Partner links channel and left it to whatever the default rules decided. The review takes minutes, and it turns silent misclassification into a visible, fixable line in a log.

Practical suggestion: give each vocabulary row an owner, a creation date and a purpose. A tag with no owner becomes a rule nobody can explain a year later. Agencies working in a client's property should also talk to the client before consuming one of the client's limited custom groups, and deletions should be approved by a second person because they cannot be undone.

Misreadings to avoid

• The default channel group cannot be edited. What you can edit is a custom group, including one you set as primary.
• A custom group changes classification, not raw data. If the tag is missing, the rule cannot find the visit.
• Order decides the result. A visit that matches several rules goes to the first.
• Reports look back, audiences look forward. The same edit behaves differently in each.
• Custom groups are not in the BigQuery export schema and cannot be used in the Key events paths report.
• A channel group does not prove that a sponsored article caused a sale. Other touchpoints may have come before.

Frequently Asked Questions

Is the default channel group editable in GA4?

No. The Analytics help page says default channel groups cannot be edited. You can create a custom channel group and set it as the property's primary group.

What does rule order change?

The page says traffic is included in the first channel whose definition it matches, given the current order of channels. A broad rule placed before a specific one stops the specific rule from firing.

Can a custom group reclassify old data?

The page says they can be applied to reports retroactively. Editing a group that an audience uses is not applied retroactively.

What is the limit on custom channel groups?

Per the page, a standard property can have 2 and a 360 property 5, with up to 50 channels per group.

Are custom channel groups exported to BigQuery?

No. The page states that custom channel groups are not available in the BigQuery export schema.

Can a custom group repair missing UTM tags?

No. Missing UTM parameters result in (not set) values in reporting. Fix the tags first, then write the rules.

Conclusion

A custom channel group lets you see ad and sponsored content traffic in your own categories while the default group stays fixed. The rule can only catch what the tag says, so settle utm_source, utm_medium and utm_campaign first, order channels from specific to broad, and test each channel with a sample link. Reports apply the group retroactively, audiences apply edits going forward, and BigQuery does not carry it. For a festival ticketing platform, seeing sponsored editorial on its own line is possible only when tags and rules are built together.

If you are planning ad or sponsored placements on publisher sites and want measurement tags agreed from the start, browse our ad spaces.