Freelance project tracker: one board for every client deliverable

A freelance project tracker is one list of every deliverable you currently owe a client, where each line carries the client, the thing being made, the due date, the stage it is in, and whose court the ball is in. That is the whole idea. Done properly, it answers two questions in under a minute: what is due next, and what has been sitting somewhere for nine days without anyone noticing.

It is a different job from time tracking. Hours tell you how long work took and what to invoice; they say nothing about where a project stands right now. If your hours are logged but you still cannot name the client blocking you this week, the missing piece is a deliverable record, not a better timer. The time tracking workflow for multiple clients is worth having alongside this one, but it will not replace it.

Why your hours do not tell you where a project stands

A three-hour entry for a client reel looks identical whether the reel shipped on Monday or has been waiting on client notes since last Wednesday. Time entries are historical: they record effort already spent. That is what makes them useful for invoicing and useless for triage.

Deliverable tracking is forward-looking. It records what exists, what is owed, and what is blocked. The failure mode it prevents is not lost hours, it is a silent stall: a card or thread nobody is currently responsible for, which surfaces as an apology two days before a deadline.

How big does this list get? The examples in this article assume a solo freelancer running three to eight clients with roughly 12 to 25 open deliverables at any time. That range is an assumption for illustration, not a measured benchmark, and your number may be well outside it. Either way the pattern holds: small enough that a single board holds it comfortably, large enough that memory stops working.

The six fields each deliverable needs

Resist adding fields. Six is enough for the whole system, and every extra field is one more thing to fill in when you are tired. Here is the record with a worked example, using the assumption above of five clients and 17 open deliverables:

Two more assumptions in that example: you are doing the work yourself, so there is no assignee field yet, and you bill per project or retainer, so hours live in a separate record. If you subcontract, add assignee and nothing else.

One board, one card per deliverable — and the column that matters

The setup that holds up is a single board for all clients rather than one board per client. Per-client boards hide the only view you actually need: everything at once. Cards are grouped by client label or filtered by client, not separated into parallel universes. If you are starting from scratch, the board setup described in project boards is the shape to copy: columns as stages, labels for clients, filters for the daily view.

The columns are stages, and the split that does the real work is whose turn it is. Every stalled freelance project is a card whose owner was never written down. Five stages cover most freelancers:

In the five-client example above, a healthy Monday might sit at 2 not started, 6 working, 4 waiting on client, 3 final check, 2 delivered. Those counts are an illustration, not a target. But the ratios are worth reading: if “working” is over about 8 while “waiting on client” is at 1, you are the bottleneck. If “waiting on client” is at 6, your week is going to be mostly follow-ups, and you should plan around that rather than discovering it on Friday.

The honest tradeoff: one shared board gets crowded past roughly 25 to 30 live cards. At that point archive aggressively — a card with no movement in 14 days is either dead or being avoided, and both belong off the active view.

Freelance tracker setup check

Build one board that answers what is due and what is stuck, and confirm client access shows only what you intend.

Milestones and revision status without duplicating effort

The tempting mistake is a second tracker for milestones and revision rounds. Do not build it. Let the stage name carry the revision, so the card reads “Reel — client review round 2” and the stage column stays simple. One record, one place to look.

File lineage is where this pays off. In kloudboard, uploading a new version to the same file keeps the earlier ones and reviewers can switch between them, so a card that says v3 points at a file where v1 and v2 still exist. Opening the Review action beside a card's attached media file or PDF gives timestamped comments and @mentions, which means notes sit next to the moment in the video instead of in an email thread with three unrelated topics.

The failure case is re-uploading a revision as a brand new file. You lose the version chain, and six weeks later nobody can tell which cut the client actually approved. Same file, new version, every time.

What clients see, and what stays internal

There are three positions, and the right one depends on how much your clients want to look.

No access. You send a status update when a deadline moves. Cheapest to maintain, and fine if clients check in weekly at most.

A summary. You paste a short list of what shipped and what is next. Duplicated effort, but you control the framing completely.

Real access. Clients log in and see their own cards. This is the version that removes status-request emails, but it is an ongoing discipline rather than a switch you flip once. In kloudboard, keep a client on the guest tier without the “view all cards” permission and they see only the cards they are assigned to; an unassigned card is invisible to them. You can also restrict a board to a member list, and you set per-member permissions in Settings > Team. The client portal overview covers what that access includes.

If you go the third way, do the test that catches almost every mistake: invite one client, assign them one card, then log in as them or have them confirm what they see. If internal production cards or the wrong client's work are visible, the fix is permissions, not a new tool. There is a fuller walkthrough in giving clients project access without exposing internal work, and the seat and access questions are covered in project management software with a client portal.

The five-minute daily review

This is the part that keeps the tracker alive. Set a recurring five-minute slot at the same time each day, ideally at the start, and work three passes in order.

1. Minute 1 — due dates. Look only at cards due in the next three days. For each, either confirm the stage is accurate or move it. Nothing else gets touched in this pass.
2. Minutes 2–3 — the waiting-on-client column. For every card past its follow-up date, send the nudge now. This is the pass that recovers days, because client silence is the single largest source of slipped deadlines and it is invisible in a to-do list.
3. Minutes 4–5 — the orphans. Any card without a due date, without a next action, or with no movement in over a week gets one of three outcomes: give it a date, write its next action, or archive it. Do not leave it sitting as decoration.

The review should produce one or two decisions, not a status report. If it routinely takes twenty minutes, you have too many live cards, not too little time.

Once a week, add a fifteen-minute pass: reconcile delivered cards against invoices, check the month ahead for collisions where two client deadlines land in the same week, and look at whether any client has been quiet for longer than their normal cycle. A card entering a stage is one of the triggers board automations can watch, and the Templates tab in kloudboard includes follow-up and overdue patterns you can adapt rather than build from scratch. Anything with an AI step spends the workspace's AI credits like other generations, so treat it as a real cost rather than free plumbing.

Start today with the smallest useful move: write down the five deliverables you currently owe, one line each, with client, due date and whose turn it is. If two of the five are waiting on a client who has not replied this week, you have already found your afternoon.