Skip to content Skip to navigation

HP's Top 10 Trends in BI (and HIT) for 2009: #5 BI — Keep It Simple, Stupid

July 9, 2009
by Marc D. Paradis MS
| Reprints

Sometimes vendors do get it (mostly) right. Hewlett-Packard put together a brief white paper in February of this year laying out their view of Business Intelligence (BI) for 2009 (and beyond). I think that they got it largely right. Their #5 trend is a rising demand for a new performance dimension, namely, simplicity. Below is a summary of the trend, my thoughts on whether HP got it right and what the trend may mean for HIT.

HP predicts: a potentially disruptive innovation arising from several IT and BI trends, such as data warehouse appliances, outsourcing, Software-as-a-Service (SaaS), Service-Oriented Architectures (SOA), and the economic commoditization of BI.

The Verdict: Yes and no. As long as I have been in BI (a little over a decade now), BI vendors have been touting their simplicity. As long as I have been in BI, using and implementing BI tools has been anything but simple. Again, perhaps I am the jaded grognard here, but I very firmly believe there is a direct relationship between the value delivered by a BI implementation and the complexity of that BI implementation (the simplest way to think of this is as a linear relationship y = mx + b, where y is the value delivered and x is the complexity of implementation; b represents the inherent initial value of the implementation and m is the proportionality constant. Of course in real-life there is a least one other dimension, z for time and m, x and b are all likely to be functions of z). While a truly simple tool to implement and use is the holy grail of BI, I’ll take the liberty of mixing my metaphors here and say that such a tool is closer to the leprechaun’s pot of gold at the end of the rainbow. After all Galahad did find the grail demonstrating his holiness in the process, while no one ever came out the better searching for the leprechaun’s gold.

HIT Impact: Essentially nil. BI is an analytical tool first and a reporting suite second. Canned reports with simple prompts, pre-defined drill-downs and hard-coded links to other canned reports are not BI. Presuming the requirements for such reporting applications are well-spec’d, they are trivial and cheap to implement with sufficiently talented developers and engineers, presenting minimal barriers to adoption. BI in the HIT context requires subject matter, analytical and clinical expertise to collect and interpret ambiguous requirements, to derive and/or validate algorithms to implement those requirements and finally to assess the clinical relevance of any findings and to initiate necessary process improvements from those findings. Simplicity simply isn’t an option.

 

Disclaimer: The opinions expressed herein are my own personal opinions and do not represent my employer's view in any way.

Topics

Comments

A colleague of mine e-mailed me the following question:

"Conceptually I like the function of complexity/simplicity in the value prop for BI. The equation doesn't work for me though. It looks like you are saying that y (value) b (initial inherent value) plus an additional amount of value that scales to complexity (mx). So the y (value) can never be less than the b(initial inherent value)?"

And he is correct. Linear functions are a form of straightjacket. Very effective at what they do and very simple to understand what they do, but also very limited in what they do.

Of course, in reality, the value-complexity relationship is anything but linear (as noted above), but rather a complex multivariate function that includes time, changing business drivers, capability of an organization to effectively implement and then to adopt a BI application, and on and on.

However, everything else being constant and equal (again, never in reality), delivering greater value requires increasing complexity.

Marc D. Paradis MS

Marc D. Paradis, MS, is Director of Strategic Data Services in Applied Informatics and a Manager...