Are You Telling Stories With Your Data?
January 20, 2011 | Leave a Comment
What a great quote from Dr. Brene Brown at the University of Houston. You can see her TED Talk here for more about how she elevates the use of data to make change at her institution.
(Via Brand Autopsy)
Dashboards in Minutes?
September 23, 2010 | Leave a Comment
Previously, I posted on a couple of higher education vendors touting their “out of the box” dashboards. Here’s another example of messaging that should make every BI project lead or CIO cringe. “Data to Dashboards in Minutes” is the byline for the Pentaho Agile BI Tour that started recently. It sounds appealing: a half day workshop to help you “deliver successful BI solutions to users more quickly and at significantly lower cost.”
But what part of the dashboard deployment process are they talking about? I doubt the half day seminar with hands on demo includes the challenging, up front, and most important part of the process. That is, to answer questions surrounding data governance, ownership, quality and integration. I also suspect it sidesteps the contentious debate that occurs during most dashboard design and deployment efforts on what metrics, goals and visual presentation should be made for the various user personas who will access the them. In my experience, every user wants their own view and metric definitions of what matters for their job responsibilities. Dashboards are notoriously difficult to create with the right level of detail and interactivity, except at the very highest levels of summary. And, those are often the least useful for real decision making.
So, I am curious as to how comprehensive the Agile BI Methodology is with respect to the people and process side of the BI equation. Ignoring this key aspect and touting ”dashboards in minutes” or “out of the box” really does a disservice to everyone involved–the end user, IT, professional services, and BI vendor communities because it sets a false expectation of what is possible. It doesn’t tell the whole story of what resources are required to get the organization where it wants to go. That leaves BI professional service companies like ASR having to do more education on the realities that go beyond just the BI technologies. The limiting factor is often the organization’s own ability to deal with the internal transformations needed to understand and manage with data.
That is the big nut to crack.
Dashboards “Out-of-the-Box?” Pop Goes the Weasel!
August 3, 2010 | Leave a Comment
You should think twice whenever you hear vendor marketing claims that their Business Intelligence or Dashboard solutions are “out of the box.” What does this really mean? How can “out of the box” solutions really address the comprehensive and unique reporting or analytic needs of your organization? If you don’t ask some probing questions you might be in for an implementation surprise. Last week I attended a higher education conference and this issue came to the forefront as clients asked their vendor legitimate, pointed questions during a dashboard solution presentation. The veneer of the scripted demo and marketing message was quickly shattered.
Compare, for example, this SunGard Higher Education press release and Datatel’s product web page , from two of the largest solution providers in Higher Education, each promoting their similar Advancement Dashboards solutions. Both companies message using the phrase “out of the box” or “pre-configured” and focus on the time savings of the data being already integrated and the fundraising measures being already defined. But how do these solutions really handle the exact requirements of the hundreds and hundreds of clients between them? Datatel’s product page includes a screen shot that can be used to illustrate a few of the questions that one should ask to get a better picture of what is involved in an implementation:
- What is the time frame of the data displayed? As of right now or of a point in time? Or both?
- What dimension or categorical code values define the KPIs so they aggregate correctly?
- Where do the goals come from? How do you set the ranges for the speedometer/thermometer?
- How do you change a dashboard to show different analytics or reports?
- How do I manage security and access so people only see data appropriate for them?
The reality is these solutions are only templates that understand the source data model and how it is organized. That’s helpful, but not nearly the whole picture. You first have to fill in a lot of blanks before the dashboards will display anything. You need to engage in a process with the end-user stakeholders to determine the answers to the above questions (and more.) For example, the time frames for the queries need to be set to match your fundraising planning and strategy cycle. Your campaign code values are going to be different than any other institution. Your goals are unique and the target values are not likely stored in the ERP or in any database to automatically display the progress. The responsibilities of staff in your Advancement office will imply unique information needs that won’t always match nicely to the user roles defined in the solutions. Most of this involves configuring dozens and dozens of query filters, reporting tool settings and security options. But more importantly the “out of the box” data model will not be sufficient to support your institution’s unique analytic requirements. This will require customization. Understanding more about what it takes to extend the data model is essential.
The bottom line is you need to approach the implementation not as a quick fix tool, but as an iterative technology and business user partnership that probes into the real information and presentation requirements. The resulting business rules can then be appropriately encapsulated behind the scenes in the BI tool itself. Having come from the product marketing side at a higher education vendor, I know where this messaging comes from and why. It’s an attempt to simplify and make a BI implementation manageable and palatable to the buyer. They’re trying to sell a product that isn’t really a product. On the surface it looks like a good solution and demos well, but the devil is in the details. It is a combination of technology and business analysis. Having deployed numerous solutions like this, it’s not as simple as it seems. It takes a good methodology combined with strong technology and business skill sets to get right. Know what you’re getting into and what flexibility is available in your solution of choice and reap the benefits of BI!
Designing an Effective Dashboard
July 20, 2009 | Leave a Comment
Here’s a nice overview presentation of effective dashboard design. The presentation does a good job of going over the do’s and don’ts and also provides a few good resources. The author follows a fairly minimalist approach to dashboard design, but with an edge of practicality.
Some key points:
- Highlight exceptions only
- Go easy on the eye candy
- It’s okay to deviate from a best practice, but only in rare instances
- Packaged dashboard solutions rarely meet all of your needs.