The Dashboard Spy

Enterprise Dashboard Topic: Does anyone know who invented the digital dashboard?

Believe it or not, I meet a lot of people who claim to be the first to have done something in the dashboarding space. 

It’s actually not such a wild claim. 

Dashboard Spy readers know that this forum is dedicated to learning from the work of others. It is by studying, borrowing and extending prior examples of dashboards, scorecards and other business intelligence interfaces, that we advance the body of knowledge known as dashboarding, business intelligence or information visualization.

As someone who sits squarely in the “information wants to be free” camp, and an avid “borrower” of “OPD” - Other People’s Dashboards, it disturbs me when I hear people talking too seriously about how they own the rights to certain business intelligence delivery patterns.

Yes, I do understand the right to make money from proprietary systems. But I do think that this is a particularly murky area in that it is certainly not clear who came up with what first. As an avid collector (some say it’s a fetish) of dashboard examples,  I can point to examples of dashboard functionality that clearly establish date claims.

This all came up during a conversation with a Dashboard Spy reader that claimed to have invented some dashboard presentation technologies way back in the day. We spoke at length - speculating who invented the first dashboard, digital or otherwise. Our conversation eventually led to a joint examination of an online patent database. It let’s you research patent filings at no charge. It’s an interesting exercise to search patents relating to business intelligence and dashboards.

Here is a sample results list. I’ve circled one that I looked at in detail:

The one I circled was filed by none other than Microsoft. Here’s a look at the details:

(more…)

Portlets, widgets, web parts, panels, gadgets - there are many names for them, but whatever you call them, the dashboard portlet has become a common design pattern in dashboard design. Simple to understand and great for categorization, these groupings of KPIs can take a tabular or graphic format. Here is a typical portlet with a pie chart:

In today’s post, we have a special treat - I actually took the time to create a video tutorial in which I demonstrate how to code the above ”jelly bean” dashboard portlet in HTML. In this video tutorial, I slowly introduce HTML and CSS and show how to wire up a pie chart graph from a google docs spreadsheet. If you’ve ever had the slightest interest in seeing what HTML coding for dashboards is like, please watch the video. Let me know if you like it and I’ll do more.

Actually, I’ve got a whole stack of emails with questions from readers that I plan to answer through similar videos. I’ll call the series “Ask The Dashboard Spy”. Have a look at this video and stay tuned for more in the series.

First, let’s take a look at the letter the reader sent me:

Subject: Please teach me some dashboard coding

Dear Dashboard Spy,

I appreciate very much your Dashboards By Example blog of screenshots.
I’ve gone through hundreds of the examples and came away with some great
UI ideas. Thanks very much.

I’m a subject matter expert from the business side and so I find myself the
least technical person on the dashboard project team. I want to be taken
seriously by the IT guys so can you please teach me something technical.
I’d love to show off a little coding skill so that I can gain some “cred”
from the team.

Regards and Thanks Again.

Dashboard Coder Wannabe

So, here is the video I created in response.  After you watch the video, you can hit the ”more” link or scroll down to see the sample HTML code. Please note that the video is longish - about 30 minutes and covers a more topic that’s more technical than usual for this blog. It’s in response to many of you looking for more technical advice. I thought I’d start simple and tackle some very basic HTML and CSS that all dashboard project people should understand.

The real gem in this video is how the portlet consumes a google web service to show a pie chart. You can set up a google docs spreadsheet and use it for the backend and generate KPI graphs which you publish on the web and can bring up in the sample HTML portlet that I will demonstrate the simple code for.  Enjoy and please provide feedback.

(more…)

Dashboards work brilliantly as navigation devices.  So much so, as a matter of fact, that the dashboard has become a commonly used design pattern for the browsing of databases.

Usability experts will tell you that very specific navigation models exist. One very straight-forward model of navigation involves searching for something you know you want to find. The “Known Item Finding” search model assumes that the searcher knows about the thing he wants to find. To accomodate this model, dashboard designers will offer a search facility. This will be form based and usually consist of alternate complexity levels - i.e. a search box on the top of each page of the dashboard with a “go” button next to it and a link to an “advanced search” or a “power search” button.

Another search model is known as “Browsing”. In this case, the user doesn’t have a known item in mind, but will let the search process refine the search. Here, the value is in presenting categories and sample values and letting the user browse to an ultimate asset.

The “dashboard as navigation” school of thought certainly follows this latter model. The idea is to present a smorgashboard of content that allows for drill-down linking into content or across categories until the user finds what she is interested in. Dashboards facilitate this sort of learning by browsing.

Let’s look at an example of using a dashboard as a browsing navigation tool to examine a database.

The Marvel Database tracks all things having to do with Marvel Comics:

Yes, it’s kind of plain and straight-foward, but you know what? If you are a Marvel Comics fan, the categorization really speaks to you. One look at the portlets and you know that you are among friends. The stage is set for browsing the data in terms that you understand. Take a look at the next dashboard screenshot to see something a little more visually interesting:

(more…)

Dashboards are most effective when users can set their own preferences and choose their own content. Many great dashboards don’t allow for this at the user level, so I wouldn’t say that you absolutely must offer that feature to guarantee user acceptance of your dashboard, but it really is a great feature.

From the IT department’s point of view, offering user configuration and personalization does makes things a bit more challenging with the need to persist the latest user settings and such, but the benefit versus cost analysis is often quite high.

From a design perspective, let’s take a look at how to let users configure their dashboards. There are a few different paths to take in terms of how to intuitively design this. I’ve seen text-only configuration screens and wizard-driven dialogs. I’ve also seen in-place editing of the various dashboard portlets.

Let’s take a look at an approach where you look at a representation of the dashboard layout and select your report criteria for each dashboard portlet. Take a look at this screenshot:

In the next screenshot, you’ll see the results of your selections.

(more…)

Dashboard design is often focused on the visual aspects of the project - lots of attention is paid to the selection of graphics and charts, the right colors for the KPI indicator lights, what style the portlets should be, etc. We’ve all been consumed with getting the right level of eye candy correct.

Countless arguments between the business intelligence experts and the graphic artists have occured as to how to style charts for maximum understanding, while satisfying the need for some dashboarding bells and whistles.

Today we focus on a more fundamental challenge of business dashboard design - what do the dashboard users want out of the system?

While digging through my countless links to dashboard whitepapers I came across a chart discussing the questions dashboard users ask:

(more…)