How Networks Group Content the Way Their Business Actually Runs

YouTube organises content the way viewers browse it. Networks need to organise content the way finance, sales, and rights teams actually think about it.

Those are two different jobs, and YouTube only does the first.

The platform gives you categories, tags, hashtags, playlists, and analytics content groups. All of them help viewers find the right video to watch next. None of them help your finance director answer "what did sponsored content earn us this quarter?" None help your sales team pull "all Sky-funded co-productions" for the partner review. None help legal flag "every video with rights expiring inside 90 days" for renewal.

What is missing is a custom fields layer that groups content by the dimensions your business actually uses. Sponsor. Series. IP source. Commissioner. Rights category. Territory. Language. Production source. These are the cuts that drive real reporting, and they do not exist natively in YouTube Studio.

This guide covers what YouTube's native categorisation does and does not deliver, what custom fields mean in the CMS world, the four jobs a custom fields layer actually does, industry-specific examples, and what an effective solution looks like for networks operating at scale.

‍

What YouTube's native categorisation actually does (and what it cannot)

Before deciding what to build, it is worth being precise about what already exists.

YouTube Studio gives you five categorisation surfaces. Each one is genuinely useful for what it was designed for, and none of them solve the business reporting problem.

Categories The 15 fixed buckets you choose per video at upload: Entertainment, Education, Sports, News and Politics, Gaming, Music, and so on. They feed YouTube's recommendation engine and surface videos on the Explore destination pages. They are not editable beyond the fixed list, they do not roll up across channels in any business-meaningful way, and they were never designed to split content by your sponsor, your commissioner, or your rights structure.

Tags User-defined keywords you add at upload. They had real algorithmic weight pre-2020. They have very limited weight now. YouTube's own documentation confirms that title, thumbnail, and description carry far more discovery weight than tags. Tags are still searchable inside Studio, so they can serve as informal markers, but they are a creator field, not an operational metadata layer.

Hashtags Hashtags appear in descriptions and titles. Up to three can render above the video title on the watch page. They help with discovery surfaces and trending pages. They have no business reporting application at all.

Playlists Viewer-facing collections of videos, useful for series, themed groupings, and viewer navigation. They are not portfolio reporting. They do not roll up revenue. They do not track rights status. They are a discovery tool dressed as a taxonomy.

Content Groups (in Analytics Advanced Mode) The closest YouTube gets to portfolio segmentation. You can group up to 500 videos, playlists, or channels together and view their combined performance. This is genuinely useful for ad-hoc analysis: "how did the cricket highlights perform across all five sports channels?" can be answered with a content group. But the limitations are real. Content groups are capped, manual to set up, analytics-only (they do not write back to video metadata), and tied to a single channel context unless you build them carefully across managed brand accounts.

The pattern across all five is the same. YouTube's native categorisation is built for viewer discovery and platform-level analytics, not for the way a business actually segments and reports on its content portfolio.

‍

What custom fields means in YouTube CMS (and why most networks cannot use it)

There is one layer inside YouTube where business-relevant custom metadata genuinely exists. It is gated behind CMS access.

YouTube's Content Management System gives partners structured asset metadata far richer than anything in Studio. According to YouTube's Help documentation, CMS partners can attach detailed metadata to every asset they own, including ownership territories, asset type, contractual fields, reference IDs, and bespoke labels. Music partners use ISRC codes for sound recordings. Video rights holders use ISAN identifiers and contractual schemas covering ownership shares, territorial rights, and match policies.

This is the closest YouTube comes to a business taxonomy layer. CMS reporting groups data by these fields. Revenue can be sliced by territory, by asset type, by rights category. It is the system that lets a music label answer "what did this album generate across YouTube globally last quarter" without joining six different exports together.

The problem for most network operators is access. CMS is invitation-only and reserved for multi-channel networks, major media companies, music labels, and large rights holders with significant copyrighted catalogues. Networks managing 5 to 50 channels are almost never granted CMS regardless of revenue. The eligibility criteria centre on rights ownership at scale, not on network size or output volume.

For everyone outside CMS, the custom fields layer has to be built outside the platform. That does not make it impossible. It just means the work happens in your reporting stack rather than inside YouTube.

‍

The four jobs a custom fields layer actually does

If you are going to build a categorisation layer that sits above YouTube, it is worth being clear about what it is actually for. Four jobs matter more than anything else.

The first is revenue attribution. Finance needs to roll up earnings by dimensions that mean something to the business. Sponsor performance. Series profitability. IP source contribution, owned versus licensed versus co-produced. Sales channel split, direct sold versus network sold versus sponsored. Without these dimensions, the revenue line in your management accounts is a single number with no story behind it. With them, finance can answer questions the board is going to ask.

The second is rights governance. Legal and compliance need to know which content has expiring licences, which carries UGC-derivative material, which is geo-restricted, which uses talent on fixed-term contracts. A custom fields layer flags this content automatically when rights are approaching expiry, when a contract date is hit, when geo-restrictions change. Without it, rights tracking lives in someone's spreadsheet and falls behind the moment that person is on holiday.

The third is sales performance reporting. When a sponsor wants the quarterly performance review of their campaign, you need to pull every video tagged to that sponsor across every channel in your network, with combined view counts, watch time, geography, and revenue. The same goes for commissioner reports, partner reviews, and any other relationship where someone outside the business is asking "how did our content do?" Without a custom fields layer, this report takes a finance analyst two days to assemble manually. With one, it is a scheduled CSV.

The fourth is audit and compliance. Auditors want to trace every video to its production source, its commissioner, its rights category. They want documentary evidence that revenue has been correctly attributed across the network. They want assurance that content has been correctly classified for advertiser-friendly status and age-appropriateness. None of this lives in YouTube Studio. All of it needs to live somewhere, and a custom fields layer is where.

These four jobs are why operators end up building this layer even when they could technically live without it. Each one is a question the business genuinely asks, and answering each one manually from YouTube Studio is expensive in time, fragile under audit, and impossible to scale past a certain channel count.

‍

Industry-specific worked examples

The same principles look different in practice depending on what you make and who you sell to.

For a broadcaster, the useful custom fields are usually genre, IP source (owned, licensed, co-produced), commissioner, territory, and rights expiry. The business question this answers: "show me revenue from co-produced content commissioned by Channel 4 in the UK with rights expiring in 2027." Native YouTube fields can answer none of that. A custom fields layer keyed to video IDs answers it as a single query. The same structure also drives the contractual reporting back to commissioners and the rights renewal pipeline that legal needs to track.

For a music label, the dimensions are artist, album, ISRC, featured territory, and release campaign. The business question: "what did the new artist campaign generate across YouTube globally in its launch quarter, split by territory and ad versus subscription revenue?" Music labels with CMS access answer this natively through asset metadata. Music labels without CMS access need an external custom fields layer joining Data API video lists with their internal release catalogue.

For a sports rights holder, the relevant fields are competition, season, match date, rights window, and territory. The business question: "show me viewership and revenue for Premier League content uploaded between August 2025 and May 2026, with rights expiring 30 June 2026, in territories where we still hold distribution rights." The combination of time-bounded rights, territorial restrictions, and competition-level reporting is exactly where YouTube's native fields offer nothing. The reporting layer has to manage this externally.

For a kids and family network, the dimensions are age rating, language, series, character IP, and compliance flags. The business question: "show me all Spanish-language episodes of Series X targeting under-eights, with any compliance flags for review before the regional regulator audit." Kids content has extra regulatory weight, and the compliance trail matters more here than in most other verticals. Custom fields are how you build that trail.

Four different industries. Four different field structures. Same underlying need: business taxonomy YouTube does not provide, layered on top of YouTube's video IDs through an external system.

‍

What breaks without custom fields

The cost of not having this layer becomes visible the moment a senior leader asks a question.

Finance cannot answer "what did sponsored content earn us this quarter?" because sponsorship is not a native YouTube field. The closest Studio reporting gets is revenue by traffic source or revenue by video, neither of which separates the sponsored content from the editorial. A finance analyst can build the report manually by cross-referencing a list of sponsored video IDs against revenue exports, but that is a two-day job every quarter, and the list of sponsored video IDs has to live somewhere reliable in the first place.

Sales cannot pull "all Sky-funded co-pros performance" for the partner review without first knowing which videos are Sky-funded. That information lives in production records, commissioning contracts, or someone's email, not in YouTube. The sales team either has to maintain a parallel list of which videos belong to which funder, which goes stale within months, or they have to ask production every time a partner review is due.

Legal cannot surface "all content with expiring rights in 90 days" because rights data lives in contracts and licence agreements, not in YouTube metadata. Without a custom fields layer that connects video IDs to rights expiry dates, legal has to scan contracts manually before every quarterly review, hoping nothing slipped through.

Compliance cannot flag "all UGC-derivative content" for review because UGC status is not a YouTube field. It is a decision someone made at upload time about whether the content uses or derives from user submissions, and that decision either gets recorded externally or it gets forgotten. Forgotten is the default.

The pattern across all four is that the business question requires joining YouTube data to information that lives outside YouTube. Custom fields are how that join happens. Without them, every report is a manual reconciliation exercise.

‍

The workaround options for networks without CMS

There are three honest paths if you cannot get CMS and need this layer anyway. They scale differently, and the right choice depends on your channel count and the maturity of your data operation.

The simplest path is spreadsheet tagging. Export your video list from Studio or via the Data API, add columns for your business dimensions, and maintain it manually. Total upfront cost: nothing. Total ongoing cost: a person checking the spreadsheet every week and updating new uploads. This works for a few channels and a few hundred videos. It breaks past that because the spreadsheet drifts out of sync with Studio every time someone changes a title or unlists a video, and there is no automatic check that catches the drift.

The middle path is a custom database keyed to video IDs. The database holds your custom fields. YouTube remains the publishing system. A scheduled sync, daily or weekly, refreshes the video list from the Data API and flags new uploads needing classification. This handles 25 channels comfortably and 100 channels with discipline. The cost is database setup, sync engineering, and someone owning the data hygiene. Most operators land here when the spreadsheet breaks and they are not ready to commit to a full API-based stack.

The strongest path is an API-based metadata layer. The Data API pulls video lists on a schedule, the Analytics API pulls performance, the AdSense Reporting API pulls revenue, and all three are joined against an external custom fields table in your reporting stack. Outputs are scheduled CSVs and dashboard refreshes for finance, sales, compliance, and the channel ops team. This is the architecture covered in detail in our multi-channel YouTube reporting without a CMS guide. Total build cost in-house is typically £50,000 to £200,000 plus ongoing maintenance. Agency partnership compresses that to a per-month fee with the infrastructure already in place.

Most networks operating at 25 or more channels end up needing the third option. The spreadsheet breaks early, the database holds for a while, and then the reporting volume crosses the point where manual sync becomes the bottleneck.

‍

Building the custom fields layer that actually works

If you are building or replacing this layer, three things matter more than anything else.

The first is configuring custom dimensions to match how the business actually reports. Finance does not think in YouTube categories. Sales does not think in tags. The fields you build have to come from how the business already structures its reporting. Sponsor. Show. Series. Strand. Commissioner. Funder. IP source. Each of these needs to be a queryable dimension keyed to video IDs, not an afterthought added later.

The second is scheduled synchronisation between YouTube data and your custom fields table. New uploads arrive on YouTube every day. Titles get edited. Videos get unlisted. Channels get added to the network. The custom fields layer has to track all of that without drifting. A daily sync that flags new videos awaiting classification is the operational minimum. Weekly is too slow.

The third is reporting that joins both data sources and lands in formats the relevant teams use. Finance wants CSVs. Sales wants partner-review decks. Compliance wants audit-trail exports. Channel managers want operational dashboards. The custom fields layer is not useful until it surfaces in the artefacts each role actually opens.

At The Polar Bears, the categorisation layer in our stack is powered by Vixxi, the platform we use to consolidate YouTube, Google Ads, and Google Ad Manager into one workflow. For the broadcaster, publisher, and media company clients we work with, it handles the custom dimensions, the scheduled sync, and the role-specific reporting that Studio alone cannot deliver. The tool itself matters less than the principle. What matters is that finance, sales, and rights teams can ask their actual questions and get answers in the format they need.

‍

FAQ

What are YouTube content categories?‍

YouTube content categories are 15 fixed buckets that creators select when uploading a video, including Entertainment, Music, Sports, News and Politics, Education, Gaming, How-to and Style, Comedy, People and Blogs, Science and Technology, Film and Animation, Pets and Animals, Travel and Events, Autos and Vehicles, and Nonprofits and Activism. The category helps YouTube's recommendation system understand what kind of content the video is and which Explore destination pages to surface it on. Categories are not customisable beyond this fixed list and were not designed to support business-level segmentation.

‍

How does YouTube categorise videos?‍

YouTube uses a combination of the category you select at upload, the title, description, tags, transcript, and viewer behaviour to classify content. The creator-selected category is the starting point, but YouTube's machine learning systems weight the actual content signals more heavily than the manual category choice for purposes of recommendation. For viewer discovery, this works well. For business reporting, the categorisation is platform-facing and does not address how organisations actually need to segment their portfolio.

‍

Can you create custom categories on YouTube?‍

Not natively at the channel or video level. The 15 platform categories are fixed and not editable. The closest YouTube native equivalent is Content Groups in YouTube Studio Analytics Advanced Mode, which let you group up to 500 videos, playlists, or channels together for combined performance analysis. Content Groups are useful for ad-hoc analysis but are capped, manual to maintain, and only available in analytics views. For true custom categorisation across a business portfolio, the layer has to be built outside the platform through an API-based metadata system.

‍

What is the difference between tags and categories on YouTube?‍

Categories are a fixed list of 15 platform-defined classifications, selectable per video at upload, used by YouTube's recommendation system and Explore pages. Tags are user-defined keywords added by the creator. Tags had significant algorithmic importance before 2020 but their weight has reduced substantially. Categories still inform algorithmic recommendation routes. For operational segmentation across a portfolio, neither offers what a business taxonomy needs.

‍

What are content groups in YouTube Analytics?

Content Groups are a YouTube Studio Analytics Advanced Mode feature that lets you combine up to 500 videos, playlists, or channels into a single performance group. Once created, the combined views, watch time, revenue, and audience data can be analysed together. Content Groups are useful for cross-cutting analysis but they are manual to maintain, capped at 500 items, analytics-only, and tied to single-channel contexts unless built carefully across managed accounts. They are the closest YouTube native feature to portfolio segmentation, but they do not replace a true custom fields layer.

‍

How do you organise multiple YouTube channels?‍

For viewer-facing organisation, playlists, channel sections, hashtags, and categories cover the basics within each channel. For operational organisation across multiple channels, the practical approach is to maintain an external custom fields layer keyed to YouTube video IDs and channel IDs. This layer holds the business dimensions that YouTube's native fields do not support, syncs with YouTube via the Data API, and feeds your reporting stack. For networks with CMS access, asset metadata inside Content Manager does this job natively. For networks without CMS, the custom fields layer is built externally either in-house or via an agency reporting infrastructure.

‍

Want content reporting that matches how your business actually thinks?