[Certain liberties have been taken to illustrate a point, and does not represent any of my corporate affiliations nor employment parties. Statements are intended to generalize particular workplace concepts in an effort to reach a broad network of professional and interest readers, and should always remain classified as a ‘focus group one”s opinion – my own]
Business Intelligence is a subject I think of so often, that I fall prey to believing that all business-minded folk are partaking of the same general thought – And while there some out there who are, (quick quiz to know if you are: Do you/Have you ever done/said ANY of the following:
a) I love the new Excel 2007/Office 2007,
b) I’m a data geek [women – you know who you are and how we use this line when we have a crush on someone in the technical realm of your workplace, especially as you refer to your quirky interest in a particular technology, software product or game ]
Aah, but I degress …
or c) have ever been excited when selected to be a part of what feels like "the exclusive BETA club", a particular scenario that yields you, the end user, early exposure to a previously unreleased and new piece of software, often times involving scenarios where a technology company actually wants and solicits your feedback about what things they could change, add or remove before that product "RTMs" (releases to market).
If you answered ‘Yes’ to any of the above, you are not the folks referred to above, you, who have a love so deep for business intelligence, or at least if they do not love it, they hace a deep knowledge begrudingly earned over the years earning them a prosperous but unrewarding career. No, not you. It is the ones who haven’t a clue what BI is, not knowing that most people in the workplace are using, or have used/interfaces with a spreadsheet software, and that act, in essense, is BI. It is more than a fancy set of terms that when strung together form a very intelligent phrase – ‘Business Intelligence’ sounds smart…Right?
Well, not to those who answered ‘Yes’ to the above quiz, like myself.
But, if BI is more than just your fancy, ie.you work in the field, you will know that getting the generation that they dubbed the baby boom, from the data warehousing world, to see a new concept in a space they ‘own’, is almost impossible. Likewise, the propensity for the generation dubbed X, Y, or Z to slow down to understand the principles of warehousing is pretty low. No matter, there seems to be an equal likelihood for folks to feel the sense of what I call the "this is mine" syndrome –
Often, when automation is introduced, albeitly a feeling I expereinced once in my career, this fear of ‘your taking away the core function of work I preform, so what will I do to be valuable / keep my job’ is at its highest. Employees will wrap their hands around whatever it is, consciously or not, and both try to appear as an expert and capable of producing a huge lift in execution "i.e. jump-starting" an employee. They (I) can feel that they lack the ability in the eyes of their leadership, and may or may not become challenged to do more/better, or become resigned and less productive over time. But for me, within a matter of days of the project launching, I felt incredibly connected to the new automation project requirements gathering, thus, driving how the solution to save me time in my day, would function, serve needs and deliver value.
Thinking broader about this experience brings me back to Open Source BI, and specifically, thinking about how to build a simple reporting architecture into a robust suite of a data visualization applications and methods. Think of an Excel workbook, which includes 3 worksheets by default. Each worksheet can hold tables, images, graphs, charts and other data visualization techniques. Think about an example analysis that has a chart, a table supporting the chart, supplemental text and drop down boxes with filter for report authorer consistency for reporting — this has 4 data visualization elements to it and was very specific to a certain analysis I had in mind (Time and Motion study) —
Now, consider if I was 1 of 10 analysts in a company. If all 10 of us analysts each built a report using the same number of elements from above, or 4, 10 analysts x 4 elements = 40 possible data visualization elements (exponentially increased by the combinations of any 1 of those elements). Think of how much of that could possibly be inconsistent depending on the report developer/analyst subjectivity picked up by working in one operating unit over another.
Now, think about the challenge of creating a consistent BI platform, that allows end users to create a customized report based on varying needs for data.
It’s not easy – Companies that grow rapidly often experience the pain of having ‘multiple versions of the truth’ floating around due to the many spreadsheets introduced for the same number/figure/KPI recorded. Disparate databases make for an ever greater challenge as one strives to report on information intelligence.
What if we could offer you a blank slate, tabula rosa if you will, where each element or each thing,. product or tool that you have access to in your company, became a widget, a little tool you could enable and use as part of your larger blank slate screen. You add a graph to the top left, and a chart to the top r, connecting both elements to a data source like a Business Object Universe, or a MOLAP cube, or Relational db, or flat file like another excel workbook. Who cares…depending on your company, you may have many options or a few. In essense you are connecting your one ‘Tabula Rosa" : to: many potential data sources. Then, you add a suuplemental table, a hyperlink to a printable view (of the databoard you are building on the fly) and an ‘Export to Powerpoint, Word, Excel, Crystal, .SSRS (Reporting Services,. WebI – whatever, you get the point), all in one interface, like a simple ASP webpage, called a ‘dashboard’.
Now imagine each widget was built from code (well, this is real not imagined), and that the code, while protected in a code repository like Source Safe or Perforce from permanent deletion and other code securing modes, was offered to you, your developer or consultant, for ease of access. Then, imagine we offered you a ‘sandbox’ or server for you to publish your BI elements to, thus growing our overall BI web services, content etc. Much like a wiki like Wikipedia does to collect it’s information.
No matter the technology, those services exposed could be extracted for use in this application, connectors and all. While I know it is awhile before we can make this happen, it we started thinking about it now, while offering a more iterative approach to data mart development (ie – how useful is it for a BI team to function without a decent data warehouse,. or at least, a quick response development cycle) following the standards as necessary but deliverying in project execution. Add onto of that a solid engagement model, and a robust and customizable series of dashboards.
This is merely scratching the surface of the next gen BI that gets me really exicted.