Projects rarely fail at the deadline. They fail in week three, when nobody can say what is overdue. Here is the minimum you need to track — and why one sheet per job beats a subscription for most businesses.
Four columns: the task, who owns it, when it is due, and what state it is in.
That is genuinely enough for most work. Everything beyond it — dependencies, critical paths, resource levelling, burndown charts — is useful on large projects run by people whose job is running projects, and is overhead on everything else.
The failure mode is not insufficient sophistication. It is that nobody updated whatever you have.
Not at the end. The end is where you find out.
Week one, everybody knows what they are doing because it was just discussed. Week two, mostly fine. By week three, somebody is waiting on somebody else and has not said so, two tasks have quietly slipped, and the person who could have resequenced things does not know yet.
By the time it is visible, the options have narrowed to overtime, lateness, or both.
Three specific causes account for most of it.
Ownership by committee. A task assigned to a team is assigned to nobody. One name, always, even when three people will do the work.
Silent blocking. Somebody cannot proceed and waits rather than saying. This is not laziness — it is that raising it feels like complaining. A visible "blocked" state makes it a fact about the project instead of a complaint about a colleague.
No fixed moment when it gets looked at. A tracker nobody opens is a document, not a tracker. Fifteen minutes at the same time each week is worth more than any feature.
The four columns, and a few refinements worth having.
Task, written as a completed outcome. "Materials ordered" rather than "materials". You can tell whether it is done.
Owner: one name.
Due date, as a date rather than a duration. Durations require arithmetic; dates do not.
State, from a short list: not started, in progress, blocked, done. Four is plenty. Longer lists produce debates about whether something is at sixty or seventy per cent, which is time spent on the tracker rather than the work.
Then two optional columns that earn their place. A note field, for the one line that explains a blockage or a change. And, if the work is billable, the hours actually spent — because a project tracker that also tells you what the job cost is the difference between knowing you delivered and knowing whether it was worth delivering.
One sheet per job, not one sheet for everything. Every job has its own tasks, its own people, its own dates. Merging them produces a document too big to read and impossible to hand to anybody. Separate sheets also mean you can send one to a client without exposing the rest of your business.
A fit-out with eleven tasks across four people.
Untracked, everything looks fine until week three, when it emerges that the electrical work has been waiting nine days for a decision nobody realised was outstanding. Nine days cannot be recovered. The job runs late, and the overtime that partly rescues it costs more than the margin on the whole contract.
Tracked, that task shows as blocked on day two, at the Monday review, with one line saying what it is waiting for. The decision gets made that afternoon. The project does not become a crisis, because a crisis is just a small problem that stayed invisible.
When is a spreadsheet no longer enough?
When several people need to edit it simultaneously, when you are running many projects at once with shared resources, or when dependencies genuinely drive the schedule. Below that, project software mostly adds administration.
How often should it be updated?
By owners as things change, and reviewed together once a week at a fixed time. The fixed time matters more than the frequency.
What about tasks with no clear owner?
They are not tasks yet. Either give them a name or accept they will not happen.
How do I stop it becoming a chore?
Keep it small. Eleven tasks that are current beat sixty that are stale. Tasks nobody has looked at in a month either need doing or need deleting.
Should the client see it?
Often yes, for the parts that concern them. A client who can see progress asks for updates far less often, and a client who can see that they are the blockage frequently resolves it without being asked.
The KISS Project Tracker is a spreadsheet with one sheet per job — tasks, owners, deadlines, status, colour-coded and sortable. No subscription, no accounts to create, no onboarding. Open it and use it. Five euro, one file.
See the templates →