Calling Edward Tufte, when you get done fixing Recovery.Gov can you stop by at the Patent Office to fix their new Data Visualization Center and the Patents Dashboard? They really need an information visualization makeover if anyone is going to learn anything useful from this data.
Yesterday the Patent and Trademark Office released its new Data Visualization Center Patents Dashboard. (Here's the link in case USPTO website isn't being friendly to links. http://www.uspto.gov/dashboards/patents/main.dashxml.) The Commissioner's blog, Director's Forum: David Kappos' Public Blog, says, "This tool will give the public access to traditional measures of pendency as well as several new pendency tracking measures. We are also providing other important data covering USPTO patent operations in a convenient dashboard format." (Emphasis added...)
The most important message in the Director's post is that "...the patent backlog is stalling the delivery of innovative goods and services to the market while hindering economic growth and job creation....the current backlog of more than 700,000 applications may cost the US economy billions of dollars due to what is called "foregone innovation" -- commercial opportunities that fail to get off the ground because of long delays in necessary patent protection." In short, people simply give up and move on because they don't have the time or energy to wait around for USPTO to look at their applications. Things don't get built, companies don't get started, people don't get hired.
Ok, so what does this new Patents Dashboard do to assist the public at large and the IP community in particular in understanding what's going on at USPTO? Not much.
USPTO's new Patents Dashboard obfuscates, renders obscure, unclear or unintelligible, the urgent and critical patent pendency issues this data reveals. This unsophisticated and almost content free presentation of data does a disservice to the Examiner Corps and the inventors and tax payers who pay for all this and rely upon USPTO's decisions. The cartoon-like speedometers don't convey to the viewer either the complexity of issues that drive pendency or the economic urgency of solving the pendency problems. Little morsels of data are buried among animated speedometers objects complete with moving dials in an array of colors that have no meaning other than to offset the black background of the screen. So much screen real estate, so little data. Or as our Marine pals would say, very bad signal to noise ratios. So much for transparency.
The dial above shows that Traditional Total Pendency in months is 35.4 months. That's almost 3 years. This is an average. There are plenty of Art Units where the pendency is considerably longer. The speedometer is significantly larger than the data it presents. The data being conveyed is smaller than the dial's black background. But what's the context? What does this mean? When you scroll down to the text that accompanies the colorful dial what you learn is that pendency is actually going up for the last five months. Clearly the dial is moving in the wrong direction but who can tell by looking at it. This is alarming.
The choice of green to show the pendency information conveys the "all systems go - we're at the green light" message. The color selection has absolutely no meaning. What is the reason that the Fiscal Year Production speedometer is purple or that the Backlog is in Orange? Anyone who is serious about building a dashboard knows that color is an important visual cue to let users know what's going on. What's going on here is the use of color to distract from the data.
There is no way to look at the dial and understand how this measures up against previous performance or the goals USPTO has set as part of its strategic plan. The accompanying text for the Traditional Total Pendency speedometer says, "Our goal is to reduce Traditional Total Pendency to an average of 20 months by 2015." Where are these metrics incorporated into the presentation? What about some kind of visual indication of where the pendency is going relative to this goal? How about one of those Tufte sparklines that shows the history over a longer timeline to give the reader some indication of how USPTO is going to cut 15.4 months off the current pendency in just over 4 years? Here's a Back to the Future Question: When was the last time USPTO had a 20 month pendency anyway?
Then there is the whole issue of whether the visual representation matches the data. Here is a metaphor where improvements mean that the speedometer is going to look like its going slower while the bar charts that accompany them are going to show values declining. This is intuitively and cognitively hard to process. How does this give the viewer any sense of when things are improving? A very odd approach at information transfer.
And then there's the Patent Examiner dial. All in white except for the graphics that are all in black. Are these guys unworthy of a color or is the goal to present the grime statistics in black and white?
Here's what's troubling about this dial and the write up that accompanies it - there is absolutely no linkage between pendency and the number of patent examiners. Isn't the number of patent examiners a major issue for USPTO? USPTO recently issued a call to industry to have retired Examiners return and for skilled patent professionals to consider a career as an examiner. How does this link with the data they are presenting here? Doesn't this chart indicate that Examiners are bailing? How does this number match up against attrition? If every month USPTO adds 50 examiners but 75 leave, things are not good. The number of Examiners by itself doesn't indicate what's going on. How hard would it have been to have an indicator for the target number of examiners and measure the current count against the target. Let the user know what this number means vs. the goal.
Doesn't it make sense to show the number of examiners relative to the Technology Centers where the backlog is? How deep is the backlog relative to the workload of the examiners who are expert in the subject matter of the patent application? How does this shortage of Examiners reflect the availability of similar talent in the US? Are we running out of skilled chemical engineers or are the chemical engineers choosing to work somewhere else? Is the pendency tied to the Examiners who are available to examine the patents or is it tied to the complexity of the patent applications being submitted? How about the average number of claims by Art Unit? What about the average number of patent and non-patent prior art references? We know that Life Sciences patents are longer and have more unique words and concepts that the "average" patent (whatever that is these days.) We know that fear of being charged with inequitable conduct is causing some applicants to throw in the every conceivable reference. How does this behavior impact pendency? But can you learn anything about what's going on here from this "Dashboard"?
As they used to say in the old Wendy's commercials, "Where's the Beef"?
Aside from the typos in the content, the downloads that don't work, the ridiculous gauge metaphor, all of which is covered under the "this is a beta release" disclaimer, how much did this thing cost to build? Is this yet another part of the squandered IT budget? Did someone go out and buy a widget tool to put this thing together? How long has this been in the works? Who is supporting it - USPTO personnel or one of their contractors that just got slammed by OMB IT Dashboard? PTO might take a few lessons on effective use of dials on a dashboard by having a look at the OMB IT Dashboard.
Edward Tufte is teaching his one day Presenting Data and Information course in Arlington at the end of the month. USPTO's money would be well spent sending a few of their data visualization people to the course and then build something that conveys information rather than cartoon dials.
Perhaps Dr. Tufte can swing by and provide some guidance after the course is over.