Planning: timelines, roadmaps and Gantt charts

The criticisms of planning are always the same: plans are never realistic, they limit creativity, they constrain, they are out of date the next day. There is some truth in that. But successful projects don’t depend on luck: they depend on choosing the right technique for that project, and keeping it up to date. A plan is not a promise about the future: it is the fastest way to notice that something doesn’t add up, while there is still time to fix it.

Timelines, roadmaps and Gantt charts are not synonyms

All three tell the story of a project over time, but to different people and in different detail.

Tool	What it shows	For whom
Timeline	The main milestones, in sequence, with their dates	Anyone who needs the big picture
Roadmap	The direction: objectives, work areas, releases, by quarter or year	Whoever sets the strategy
Gantt chart	Every task with its duration, owner and dependencies	The delivery team and the PM

The timeline is used at the start and to communicate; the roadmap keeps people aligned on goals that may change; the Gantt chart is for getting the work done, day by day. Together, they tell the same project at every level.

A little history

Henry Gantt (1861–1919) was an American engineer who worked on the organisation of production. The charts that bear his name were made famous by Wallace Clark, in 1923, with a book that described them as a working tool of management: a simple way to plan work, to present facts so they are easily understood, to eliminate idleness and waste, to get things done on time. A century later, the definition still holds.

How to read a Gantt chart

The horizontal axis is time; the vertical one lists the tasks, grouped by phase; the bars show their duration and sequence. Each task has a name, a duration, a planned and actual start date, assigned resources, a place in the hierarchy, a percentage complete.

What to look at, in order:

1. The tasks, starting from the top-level ones: what has to be done and how.
2. The milestones: the essential dates, such as deliveries to the client, approvals, materials arriving. They are checkpoints, so they have zero duration.
3. The dependencies: what happens if a task finishes late, and which others it drags along.
4. The resources: how many people are needed, and when.

Two warnings. There are often two versions of the same Gantt chart, an internal one and one for the client: always know which one you are reading. And if something doesn’t add up, say so. It sounds obvious; it isn’t.

Effort and elapsed time

This is the distinction that confuses people most. Effort is the work actually needed; elapsed time is the time that passes between start and finish. They often diverge:

• two days of work given to someone allocated at 50% become four calendar days;
• two days of work starting on a Friday finish on Monday evening.

The client’s Gantt chart is based on elapsed time, because it must give dates you can keep; the internal one on effort, because it is used to allocate people.

How to build one, step by step

Before opening the tool, three questions: who will read this Gantt chart? Do I know all the top-level tasks? Is there a date that cannot move?

1. The top-level tasks. No dates, durations or resources. The Gantt chart is a document that evolves from the very start.
2. The detailed tasks. Inside each container, the individual tasks. The ideal starting point is the Statement of Work and the cost estimate.
3. The client’s steps. Sharing, reviews, approvals. These are the hardest tasks to estimate, especially with a new client: when in doubt, keep them generous in the first draft.
4. The milestones. The dates to share with client and team.
5. The dependencies. All in sequence at first, without worrying about the end date. You optimise later, on solid ground.
6. The resources. A basic allocation, which also helps check that the sequence is right.
7. The working days, from the estimate.

Then the details that make the difference: tools know the working week, but not always the public holidays, which have to be added by hand. And times of day: “delivery on day X” must also say by what time. Finally the check: if dependencies and allocations are right, the resource usage view shows no overloads.

When the deadline doesn’t add up

It almost always happens: with every task in sequence, the end date falls after the deadline. There are three ways to recover.

• Run tasks in parallel with more people. The Gantt chart lets you simulate: how many people do you really need to stay on time? Before sharing the solution, though, check that they are actually available.
• Review the dependencies. Are there tasks that can start without the design being approved? Can the back end start before the front end? Are there independent branches of work, such as two apps for two operating systems?
• Review the client’s timing. An approach that works: first show the Gantt chart without optimisations, with the date it produces, then the version that meets the deadline. It makes clear that fast feedback and approvals are a commitment on both sides. Sometimes the client won’t be able to speed up; but having made it clear from the start prevents many problems later.

An exercise

A microsite and a social landing page for the launch of a new handbag collection. Sold: UX 3 days, concept 5, design 6; microsite front end 15, landing page front end 6, back end 8 (the client wants to manage part of the content themselves); testing 3, release 6. The client asks for one week to review the design. Kick-off on 30 April, commercial launch on 24 June. Does it fit? What can run in parallel?

The plan is alive

A Gantt chart is not a document to sign and file away: it is a tool to update, every week, with what actually happened. Can you make a Gantt chart in a spreadsheet? You can draw something that looks like one; but without dependencies, calendars and resources it isn’t a Gantt chart, it is a drawing.

And when a project is too uncertain to be planned in full at the start, planning doesn’t disappear: it gets shorter. That is what agile methods do, with short cycles and a list of priorities reviewed at every round.