At Limina we spend a lot of time working on complex information display where users are making critical decisions based off of key content, often displayed in tables.  Therefore, it is very important that the formatting of tables be as consistent and clear as possible. Tables should clarify and enhance the information they present, not obscure it.

I like to think that most of us involved in GUI design (be it on the interface design or development end) have either read Tufte’s books or heard about his useful guidelines for laying out and designing tabular data. Yet, one of the most common issues we run into when performing Expert UI Evaluations is poor table design.

Googling around the web I found few sites discussing the topic of table design other than those providing libraries of CSS table designs; however in many cases these reference libraries actually add to the proliferation of bad table design. Developers get excited when they figure out a way to code up a new table style or interaction method and post it for others to use. But there is little discussion on where and when (if ever) these solutions are appropriate. Remember, just because you can doesn’t mean you should.

Here are the three areas where we commonly see the most UI design and usability mistakes and some guidelines that should help you create more useful and usable tables:

Design Style:

Help users focus on what they can learn from table data; the data should stand out, not the design of the table.

Minimize visual clutter and avoid over styling

• Do you really need horizontal and vertical grid lines and an alternating row color?
• I can’t think of any good reason to ever use a patterned background

Avoid over use of color

• A basic table shell does not need more than 2 or 3 colors to differentiate column headers, a content/grid and a selection highlight
• When you use colors use them consistently. Users need to know they can count on the visual cues you provide them so they can act efficiently
• When using an alternating line color keep the contrast to a minimum and use soft colors that are easy on the eye
• Don’t make the alternating line color too similar to your line item selection color

Content alignment (left, right, center)

• Know your content alignment rules. Typically content is aligned left. But numerical content can only be revealing if the column is appropriately aligned and left alignment is not usually the best choice.


SOURCE: http://blog.editage.com/?q=Aligning-Columns

Use cell padding and spacing – cramped table cells are much harder to read, make sure you provide enough space to allow users to easily scan content

Interaction style:

Unless explicitly called out, a user will typically need to spend time “discovering” interaction features.  In the case of tables, there are a number of visual cues you can provide to direct their interactions.

• Define a set of table styles that lets users know how they can interact with table content; read-only, editable, selectable, etc
• Apply the styles consistently
• Consider revealing controls on-hover to decrease visual clutter for scanning, but enable the table for interaction when necessary
• When designing complex tables that have mixed content types (e.g. read only and editable and/or selectable) provide a visual style that illustrates these key differences
• Clearly indicate which columns are sortable and which is the default
• Indicate whether users can resize columns and rows
• Consider using mouse over text to display lengthy cell data when truncating
• Avoid using fixed tables inside panes or windows that horizontally resize. The reason for this is because it will be easy for users to accidentally lose some columns by resizing a table-containing pane to be horizontally smaller

Table usage:

Is a table the best format for displaying your content? Sometimes viewing data in a tabular format doesn’t help your users see trends in the data. Ask yourself if the information would be better displayed using a simple (or sometimes complex) visualization. Or, consider providing both a tabular and visualization view of the content.

This list of considerations is not meant to be the definitive guide for good table design but it should help you avoid the most obvious pitfalls and put you on the right track.

-Maria

If you’ve gone shopping for orange juice anytime in 2009, you’ve been privy to one of the biggest branding blunders since New Coke. Pepsico released it’s new Tropicana brand to the market in January 2009 and shortly after, the public unleashed a firestorm of criticism over the new design.  On February 23rd, just under 2 months later, Pepsi announced they will be reverting back to the original branding.

While, for Pepsi, this has been an unfortunate loss of marketing spend and brand equity, there are some excellent lessons to take away from the experiment.

Lesson 1) Know thy consumer

At Limina we often tout the benefits of conducting user research and usability tests before, during and after the design phase of any project.  As we now know, this applies not only to pure user interfaces and product design, but to something as seemingly minor as a branding exercise.  Focus groups, as opposed to usability studies tend to remove the product and the user from context of use, and has the tendency to lose out on the rich data collection opportunities users will encounter in a live scenario.  Had the brand design team tested the new packaging in a grocery store and observed product selection behavior  of consumers with the new brand against a control study (of the original brand), they would have learned an incredible amount of qualitative and quantitative data about the impact of the new design.

My own personal observations and experience with the new design at the point of purchase consisted of stalled out consumers standing in front of the tropicana selections, heads tilted slightly  trying to figure out which one they used to buy.  On my first encounter with the new brand, I must have  spent 3 minutes crouched in front of the juice rack looking over the new and drastically reduced information based labeling of:  Some Pulp, No Pulp, Lots of Pulp, etc. etc. before grabbing the carton of choice.  Surely a brand manager would have noted this as indication of “missing the mark” and hit the drawing board for another pass before going into full blown package manufacturing.

Take Away: Test early, test often.  Context Matters

Lesson 2) Detailed  Design vs. Gestalt

The Gestalt notion,  “The whole is other than the sum of it’s parts”, could not have rung more truth in this event.
It’s one thing to look at the design of the new Tropicana carton as it sits on the table and admire some of the aesthetic choices of typography, pantones and graphics.

It’s truly another thing to put all of the variations of over a dozen cartons on the table and ask you to make a decision based on the design.  If you say you identified your preferred  household carton of choice without having to scan the selection more than once…  you’re either lying or you worked on the the brand strategy for this campaign.
Either way, running the same visual scan with the original branding is a far simpler task.  Here’s why:  The original branding had large color blocks as part of their information design system to color code a family of very similar products.  On the original design, the color coding  takes up nearly 1/4 of the visual field.  So, whether you’re a pulp, no pulp, some pulp consumer…  you knew which one was made just for you.  Looking at the new designs, the color coding system fails as it’s reduced to less than 10% of the visual field and while they added some redundant labels in case you missed the color coding, colored text on colored juice graphics don’t quite pop the way white text on color does.

Take Away: Avoid designing in a vacuum, think of how a single change ripples through the whole.

Lesson 3) Usability is for everyone

Most people who know me, but are unfamiliar with the usability field, often begin to glaze over when I start to break down what Limina does.  So I love a great anecdote that’s accessible enough for me to put people in the shoes of a frustrated user and get bobble-head nods back.  The Tropicana story is great because it’s not that they should have never touched the brand…  it’s more that they lacked an understanding of how to modify the brand without negatively impacting the information design system which worked so well.

I’ve even heard our clients say “we have a designer for that we only need you for the usability piece” or “we’d like you to run a usability study and give us the results, we can take it from there.”  It’s really not enough to see the poor ratings on a score card run with it or to take a wireframe and go nuts on the visual design.  It’s critical to see usability and information design as a holistic end-to-end process, any loss in translation of user requirements not only results in sunk cost on design/development/manufacturing, but potentially failure to take flight.  Tropicana is fortunate enough to have enough capital to absorb the blow and correct their course.

Take Away: If you’re not sure if usability applies to your product or software…  ask someone who knows something about it.

Lesson 4) Leveraging social tools

It’s no secret…  social networking tools are extremely effective in marketing feedback.  Great product managers have had their ears to the ground in social network response mechanisms for some time.  Good examples of consumer feedback mechanisms are Dell’s ideastorm.com and Starbucks’ mystarbucksidea.com, but more recently we’re seeing the strength of  Twitter and Facbook as focus group and user feedback mechanisms.  While you’re not going to get fine granularity of complex usability issues resolved in these forums, you can definitely gain perspective on broad stroke issues that ail your product.

Take Away: Word of Mouth marketing is a two way street, sure, you can promote yourself all over the social web, but take time to listen to what they’re saying in response.

Lesson 5) If it ain’t broke, don’t fix it

Needs no explanation.  Taking a page from Sony/Aiwa, Nissan/Infinity, Toyota/Lexux/Scion – if you want to test something new, create a test market or spin off a sub-brand.

I’ve heard rumors that this was a stunt to get consumer buzz.  I’m not sure I’m buying that, afterall..  they didn’t change their juice formula,  just the box design.  Whatever the case, this has been a great example of how not to mess with a good thing.

-Jon