Hi Steven,

Okay I get that you don’t appreciate BOBJ’s want to market products for sale to customers (as if all marketing weren’t some form of chutzpah)…and of course the site is new, so some of those points are aspirational…I bought your excellent book on dashboards and have occasionally developed xcelsius dashboards. What’s wrong with Xcelisus, isn’t it capable of generating bad designs or good designs. It renders data interactively, like excel charts with some extras: an excel chart can be good or bad. It’s not super-advanced but that’s better for lots of users….

Steve,

These observations may not refer to the current version (4.5), which is highly style-able include component-specific skins. For any component, you can style anywhere from minimalist to over-done. Few of those you mention are even defaults and all can be disabled. In my opinion, what happens is users go crazy, because they can, without understanding principles in your book. I myself the first thing i do is start styling, you know, just because i can. Don’t get me wrong, i am not in love with Xcelsius. It’s no matlab for tricking out and it’s a little slow to develop in, but it does allow a novice to create interactive charts out of the box.

Steve,

I don’t know if i have such examples.

I helped on a team, to produce these.
…but I helped more on the functional specs not the appearance. I don’t assume these would meet your criteria.

You know how it is, there are lots of constituents to a final beta/prototype. And, it seems to be true, that a little graphical pizzazz sometimes “sells” the information. These samples, that i linked, I know for a fact have been enthusiastically received by audiences that matter. So, you know, that’s my criteria: does it get the job done for the client, the job that the client wants done. “every product has a job.” And i absolutely do think that if a client wants snazzy graphics, then it is not for me to decide the bells and whistles are too much.A Tufte-ish chart may be better logically, but if a ringing-clanging-garish dashboard gets people otherwise interested in a boring domain, maybe it does a better job?

If you are interested, I work with, in my humble opinion, among the best dashboard producers anywhere, http://www.analysisfactory.com. Cliff and his team consistently produce client-inspiring dashboards.

David,

Unfortunately, none of the examples that you provided illustrate the ability to turn off the distracting visual effects in Xcelsius. Until I see examples of visual displays that are effectively designed for clear, accurate, meaningful, and efficient communication, I cannot upgrade my opinion of Xcelsius.

I would like to suggest that your criterion for judging the effectiveness of a data display be revised to ask “Does it get the job done for the client in the most effective manner?” The fact that a display is “enthusiastically received” does not indicate that it is effective. Many studies have demonstrated that people often prefer displays that communicate poorly. To give in to these preferences is to settle for poor software and poor performance by those who use it.

Regarding the work of the Analysis Factory, I’m sad to say that I find their dashboard designs typical and far from effective. They suffer from all of the problems that I warn against in my book Information Dashboard Design.

Hi Robin,

If I can find the time, I do plan to comment on the effectiveness of the new Google analytics dashboard. It’s better than what they provided previously, but still falls considerably short of the mark.

Steve

David,

Regarding the importance of interactivity in infovis, I absolutely agree, and so do most of the leaders in the field. Tufte’s love of paper has severely limited the scope of his influence. As the resolution of computer screens has improved, the field of infovis has extended Tufte’s insights and surpassed them by focusing on interactivity and other benefits of computer-based displays, which are critical, especially for data analysis.

Dear Mr. Bored,

My definitions of the terms “dashboard” and “faceted analytical display” are not meant to cover everything that one might need to do when displaying or analyzing data. My critique of Xcelsius and the Business Objects Insight website has nothing to do with their attempts to support what-if analysis or real-time data, which are two areas of functionality that I fully embrace. Nor do I care whether or not Xcelsius provides a bullet graph among its library of display widgets. My criticism targets the poor design of Xcelsius, which severely limits its ability to present and analyze information effectively.

As for the meaning of “transference” and the spelling of “eunuch”, you might find a dictionary helpful.

I am intentionally provocative in this blog, because being so gets people’s attention. What I say, however, is always full of substance and serves to inform. You are welcome to challenge anything that I say, and do so provocatively, but be sure to include some substance in what you say that addresses the points that I’ve made.

You are welcome to use Xcelsius to your heart’s delight. If you do, however, your work will suffer and your employer will only get a fraction of what you could contribute with a better tool. I share your frustration with “tired and remedial data analysis”, which is far too common in businesses today, but I do not share your opinion that Xcelsius goes beyond this. Like many products that dabble in data visualization, Xcelsius shows little interest in the type of data communication and analysis that businesses desperately need.

Regarding the maturity or substance of my comments, although provocative, my comments are factual and a direct response to false claims by Business Objects. When vendors make false claims or promote ineffective data visualization products, I challenge them. No one pays me to do this. I do it because I care about the business intelligence industry and about people whose jobs require them to communicate and make sense of data.

I have critiqued Xcelsius on several occasions, almost always in response to misleading marketing claims by the vendor. If you have followed this blog and my articles closely, rather than this one entry in particular, you know that I have provided detailed critiques of Xcelsius’ functionality in the past. In this particular entry I responded primarily to the vendor’s claims about the new site called Business Objects Insight, in which Xcelsius appears to play a central role.

The interactive functionality of Xcelsius, which you praise, includes few of the features that are required for effective visual communication and analysis, and the useful features that have been included are undermined by poor usability. Take a look at products that do a good job, such as those developed by Tableau and Spotfire, and you will easily see the difference between a product that plays at visual analysis and one that succeeds. My reference to Xcelsius as a “child’s toy of a product” reflects this lack of effective functionality, its focus on visual effects that have no place in business applications, as well as the fact that its developers, who no doubt did work tirelessly to create the product, came from the video game industry. Xcelsius demonstrates extensive expertise in video game development, but what it needed was expertise in information visualization.

(By the way, when I express my opinions, I use my own name and take responsibility for what I say. You might consider doing the same.)

Note: Those of you who are following this exchange regarding the merits of Xcelsius can find its continuation in my discussion forum.

Tim,

I am provoking people to work smarter and more effectively. The marketing of Business Objects that I object to, if indeed intended to be provocative, encourages practices that are ineffective. To succeed in this is hardly success.

To say that I am “academically correct” implies that I am not correct where it counts — that is in the real world. This is the opposite of what is true, however. As someone who has spent his entire professional life working in the real world of business, not as an ivory tower academic, but as a practitioner, I advocate practices that work in the real world. What I don’t do is abandon what works when people, without knowing better, ask for something that does not work. This is the difference between someone who works to do what’s right and an organization that works to sell products, whether they work or not.

Would it surprise you to hear that there are several conscientious, design-oriented employees of Business Objects who are working hard to fix the problems that I’ve identified in the products? When their efforts begin to turn up in the products, I will be the first to cheer, and do so publicly.

Bottom line – insight.businessobjects.com is just bad marketing. Content hasn’t changed in months. I’m not even sure whay they are trying to attempt – maybe they aren’t as well.