YouTube project management for multi-channel teams

YouTube project management stops being a calendar problem somewhere around your second channel. With one channel, one editor and one upload a week, a board with six columns and a publish date is enough. Add a Shorts pipeline, a second editor who cuts for both channels, and a sponsor who wants a 60-second integration inside the long-form plus two cutdowns, and three things break at once: the board's stages no longer fit every deliverable, contractors either see too much or too little, and the links between videos live in your head only.

This is the operational layer for that stage: how to structure boards across channels, how to scope contractor access when one person works on two channels, how to track a dependency between two videos, and how to run a weekly review that catches slippage before the publish date. It assumes you already have a board and a calendar. If you are still building those, start with project management for content creators for the single-channel version, and the YouTube content calendar guide for cadence.

Pick a board structure before you add the next channel

There are three structures teams actually run. The one that fits depends on how much of your work crosses channels, so start with a count: take your last 20 published pieces and mark how many touched a second channel or a paid deliverable. That ratio is the decision input.

The honest tradeoff: per-channel boards give each editor a clean pipeline but push cross-channel work into a manual step, while a single board keeps dependencies visible at the cost of a long, crowded view. Teams usually land on per-channel boards plus one campaign board once sponsor work becomes regular, because campaign deadlines are the ones with money attached.

Once boards exist, access can be restricted per board to a member list, which pairs naturally with the per-channel structure. That is a coarse control — it decides who opens the board, not which cards they see inside it, which is the next problem.

Contractor access across channels: assignment is the visibility switch

Card visibility is enforced server-side. Owners and members with the "view all cards" permission see every card. A guest without it sees only the cards they are assigned to. That single rule is what makes cross-channel isolation possible, and it also means the assignment list is your real access control.

1. In Settings > Team, pick the person and set the guest tier with "view all cards" off. Set Drive access only if they need files.
2. Set per-board access from each board's settings.
3. Assign every card they need — not just the card they work on.
4. Verify by checking that person's assignments rather than trusting the permission screen. An unassigned card is invisible to a scoped guest.

Step 3 is where multi-channel setups go wrong. A Shorts editor cutting from a long-form needs the source, and the source lives on a card assigned to somebody else. You have two options, with different side effects. Assign the long-form card to the Shorts editor, and they can open the whole card — description, comments, sponsor notes. Or share the specific Drive folder with them, and they get file access without a card appearing on their board. If their Drive access is Viewer, they can view, download and comment but not upload; that is not strictly read-only, and it is not Editor either. Sharing a folder does not upgrade Viewer to Editor.

Someone cutting for two channels needs assignments on both, which is fine, but it means their board no longer mirrors a single channel — expect that and tell them, so they stop asking why two channels' cards appear together. The full permission setup, including a test card, is in giving freelance editors access to only their assigned projects.

Cross-video dependencies: track the four links, not a Gantt chart

Multi-channel work creates four kinds of dependency. Anything else is schedule detail you can handle with dates.

• Source link: a Short cut from a long-form cannot finish before the master is locked.
• Order link: a teaser Short scheduled before the long-form, or a cutdown after it.
• Approval link: a sponsor's script approval blocks the shoot, not just the edit.
• Asset link: the final title and thumbnail depend on the locked cut, so a title change reopens both.

You do not need dependency software. You need one visible place per link. Three options, in ascending order of durability:

Name it on the card. "Blocked by: Ep. 42 master — locked cut." Free, immediate, and fragile the moment someone renames the other card.

Use a Waiting stage that means one thing. A stage called "Waiting on another video" only works if exiting it requires the specific blocking item, not just a feeling that things are fine. It also gives you a filter for the weekly review.

Automate the handoff. On a board, automations run "when X happens, do Y". Triggers include a card entering a stage, including by drag and drop, and actions include creating a new card or a task, which can go to the card's assignee. So: when the long-form card enters Locked, create the Shorts cut card or task the same day. A stage with automations shows a bolt and a count on its header; hovering lists them, and dropping a card on it shows a "Running: <name>" notice. The Runs tab records every run, and a failed run explains why with a Retry.

The limit to know: automations are board-scoped. If the long-form and the Short live on different boards, that handoff is a manual step or you keep both cards on one board until the Short is cut. Do not build a dependency trail you cannot see in one screen. For dated dependency math and slack, see video production schedule.

Same board, three deliverable types without context collapse

Context collapse is when one stage name means three jobs. "Edit" might mean cutting a 12-minute video, assembling four Shorts from existing footage, and placing a 60-second sponsor read. Reviewers then argue about whether a card is ready, because they are each measuring a different task.

The cheap fix is exit criteria, written once per format: what must be true for a card to leave Edit (locked picture, sponsor read placed, captions burned, whatever your reality is). Store it as a short document next to the board, not in a comment thread. The expensive fix is splitting the board by format. Split when the exit criteria have almost nothing in common; keep one board when they overlap by half or more.

Capacity is the other half of context collapse. More formats in flight at the same delivery rate means everything takes longer, and it is arithmetic you can do before you commit to a fourth channel.

Multi-channel pipeline cycle time

See how long a video takes from start to published at your current work in progress and delivery rate, so you can decide whether to add a channel or a person.

Average cycle time in days equals cards in progress divided by videos published per week, multiplied by 7 to convert weeks to days. This is Little's Law, which holds in a steady state: the more work in flight at the same delivery rate, the longer each item takes.

Assumptions: Assumes a steady state, that one card equals one video's production pipeline counted consistently, and that blocked time is included in the wait. It is an average, not a deadline, and says nothing about quality or revenue.

Worked example: Cards in progress across all channels: 18 cards; Videos published per week: 3 per week. Average cycle time per video: 42 days.

Worked example, using the calculator's defaults. A team has 18 cards in progress across three channels and publishes 3 videos a week. Average cycle time is 18 ÷ 3 = 6 weeks, or 42 days. Add a fourth channel with 6 more cards in flight and no new capacity: 24 ÷ 3 = 8 weeks. Hire a second editor so throughput rises to 5 a week while work in progress stays at 18, and it drops to about 25 days. The first change lengthens everything; the second shortens it. That is the decision — add a channel or add a person — stated in the same unit.

Treat the number as an average, not a deadline. It assumes a steady state, counts a card the same way every time, and folds blocked time into the wait. If your board has cards that were opened months ago and never closed, your work in progress is inflated and the average will look worse than the work feels.

Sponsor integrations: deliverables, gates and compliance across channels

A sponsor deal is a campaign with several deliverables across several channels and two sets of approvals: yours and theirs. Track it as one campaign card with a deliverable list, not as four unrelated cards on four boards.

Put the contract's requirements on the campaign card as fields or checklist items — disclosure wording, exclusivity dates, usage rights, product claims — because they are the items nobody remembers in week three. Contract terms vary and this is not legal advice; the point is that the requirement belongs on the card, not in a chat message.

One thing worth knowing about posting: board automation templates never post to social networks. The "Draft X and LinkedIn posts on Published" template drafts the posts on the card. Publishing to connected X and LinkedIn accounts exists and is meant to be confirmed with you first, so treat the draft as the deliverable and the post as a separate approval.

A 15-minute weekly cross-channel review

Once a week, in this order, with one person owning each channel. Open the boards rather than a summary document — summaries are always a week stale.

1. Waiting cards. Filter to the Waiting stage and check whether each blocking item moved. Anything waiting more than one cadence cycle gets a decision, not a note.
2. Dependencies due this week. Compare the two cards in each link: if the long-form master is not locked, the Shorts cut is not happening.
3. Sponsor gates in the next 14 days. Check the campaign card's next approval against the actual cut, not the plan.
4. Assignments. Confirm every card the scoped contractors need is assigned to them. This is the two-minute habit that prevents "I never saw it" on Thursday.
5. Overdue. If you use the "Ping the assignee when a card is overdue" automation, note that the overdue check runs hourly and fires once per due date. Assignment creates an in-app notification unless the person has turned that preference off, so do not assume a message landed elsewhere.

Escalation rule to agree in advance: a card that has sat in the same stage for more than one full cadence cycle goes to the channel owner with three options — cut scope, move the date, or add help. Without a rule, the review becomes a status meeting. More on spotting stalled work across a portfolio is in creative project tracking.

Do this today: count the cards in progress on your busiest board and the videos you published in the last four weeks. Put both numbers into the calculator above. If cycle time is longer than your upload cadence suggests it should be, you have a work-in-progress problem rather than a channel problem, and the fix is a decision you can make this week rather than another board.