What’s Wrong with Wall Charts

In February 2002, an unsuspecting project manager posted on Edward Tufte’s blog asking for help making a Gantt chart prettier. What followed was probably the most epic convergence of project management expertise we’ve ever seen — the thread is 161 posts long and still going!

The punch line? According to the dozens of assembled experts, instead of sharing around electronic copies of Gantt charts – PowerPoint slides, PNG files, etc. – project managers should make giant wall charts of their project plans to hang in the office. That way people can eyeball them, from close and afar, and visualize project status at any level of detail they desire.

I couldn’t disagree more. Here’s why.
Continue reading

Why Customers Hate Your Proposals, but not the Eiffel Tower

eiffel-towerAh, Paris in the summertime: streets jammed with tourists, peddlers pushing fake Chanel bags, and the occasional pickpocket thrown in just for good measure. Of course, there is still the Eiffel Tower, rising 986′ (301 m) into the air and reminding the masses why they came in the first place.

We all think of the Eiffel Tower as the quintessential Paris icon, but it had its share of detractors and false starts. In fact, Gustave Eiffel wasn’t a shoe-in for the design and construction at all. Eiffel worked diligently to win the favor of President Jules Grévy and the committee of the Exposition Universelle (World’s Fair) of 1889 .

Do you know how Eiffel didn’t pitch his winning design? With something like this:
Continue reading

A Brief History of Time(lines): Henry Gantt and his Revolutionary Chart

Given that we make Gantt chart software, I’ve always been a little embarrassed at how little I know about H.L. Gantt. So I dug up an old book (and I do mean old — it was published in 1922) called The Gantt Chart, a Working Tool of Management by Wallace Clark. (Fittingly, the edition I read was published by Forgotten Books.) Here’s a little historical background on the Gantt chart for all you Gantt Artists out there:

Like many great American innovations, the Gantt chart was born of war. General William Crozier, Chief of Ordinance for the U.S. Army, hired the industrial process consultant H.L. Gantt (pictured below) in 1917 to help the army prepare for the U.S. entry into World War I. A mobilization of men and munitions had not been attempted on this scale in America since the Civil War 50 years earlier, and logisticians like Crozier were scrambling to ramp up industrial production.
Continue reading

Fifty Shades of Gray in Project Reporting

A colorful project report is not always the most informative, depending on the information you are trying to communicate. Take this Gantt chart below as an example:

Gantt chart color-coded by hours of work

It’s clean and attractive, sure, but if you look at the legend in the lower-left hand corner, you will see that we are color-coding by the number of hours spent on each task. The color scheme doesn’t intuitively tell us which tasks take more work than others. Viewers have to look back and forth at the legend to get this information, which creates unnecessary whiplash on the eyeballs.

Consider this alternative:

Gantt chart with a greyscale representation of work hours

This is the exact same Gantt chart, but we’ve replaced the unhelpful rainbow theme with a grayscale representation of hours required per task. Tasks requiring the most work are the darkest, which helps us immediately identify that “Task 9” is going to take significantly more time than everything else in the project.

“Task 9” also happens to be one of the shorter-duration items in the project plan, so the grayscale approach to hours of work really helps it stand out where it wouldn’t have in the more colorful approach.

If you think gray is boring, we won’t argue with you. But luckily, you can show the relative size of different aspects of your project without feeling like you’re publishing an old newspaper. Here is an example of the same project color-coded by cost instead of hours:

Gantt chart with a green scale to represent project cost

Here, we’ve used a “greenscale” approach instead of a grayscale approach. It’s a little more lively, and since people think of green when they think of money, it’s a good choice. Again, we immediately see that a relatively miniscule task (“Task 6”) is eating up significantly more budget than everything else in the project.

So, color isn’t always bad, but it needs to be used the right way, in service of what you are actually trying to present. When building a project report, ask yourself which of these types of dimensions best applies to the way you want to assign color:

  • Categorical: These dimensions of your project create distinct categories, but doesn’t have any real order or scale. Status, resources, subprojects, and clients are good examples. Choose distinct, vibrant colors for these so that they will stand out from each other easily.
  • Interval: These dimensions of your project have clear numeric values attached, such as cost, hours, percent complete or slack. Choose grayscale or a single color with different levels of saturation to represent the high and low ends of the scale of values you are reporting.
  • Ordinal: These are less common in project management, and are typically limited to sequences (1, 2, 3). Examples include a sequence of phases or milestones. We recommend a distinct, vibrant palette (like categorical) for these so they are easy to spot, but without inadvertently implying scale.

In short, if there is some sense of scale to your project variables, you should use a grayscale or color scale to represent them. If not, you can stick to good old-fashioned color coding.

Why the Legend is NOT Doing You More Harm than Good

Often times, those of us who are responsible for upward project reporting are unfortunately limited to the usable space we are provided to communicate.  Status, risk, issues, resources, and phasing are just some of the necessary dimensions of project data that may be required (even within a single slide) in order to provide our leadership with the information they need to help make decisions, remove obstacles, or ensure alignment across a portfolio.  Inevitably, we run out of white space for text labels or clarifying bullets, and in that situation the report will require a difference in formatting to provide meaning in the most efficient way possible.  Coloring, borders, shape, and label font are all examples of differences within our chart that may have separate meaning. Instinctually when a variation in formatting exists, our eyes will search for a legend to help us understand the meaning of those distinctions and for that reason a legend must always be present if we use formatting to illustrate meaning in our reporting.  Without a legend we may be inefficient in communicating our information, or will leave room for audience to make assumptions or have questions about the information.  Instead the communication should stand on its own. Continue reading