Bestselling author Steven Covey has made a fortune with tips like begin with the end in mind. I've never met the man, but I'm virtually certain that Covey was not talking specifically about data visualization when he gave that advice. Yet, it's remarkable how often that ostensibly logical axiom seems to make its way into data-related matters.
So says Boris Evelson, a VP and principal analyst at Forrester Research. Evelson advises that organizations know the specific business metrics they want to measure before embarking on visualization projects. He continues:
Visual analysis and data visualization applications should also support the calculation and measurement of business metrics. Potential customers should have a clear sense of what metrics they will need to focus on, according to Evelson.
“For example, customer profitability by individual customer or by region, time, product line, or sales territory, and how often do these parameters change? These factors determine whether the implementation should be SQL-based, or in memory-based, with free exploration, with or without any limitation,” Evelson says.
Evelson seems like a smart enough guy and maybe he's right about certain hardware requirements. With respect to data visualization in general, though, he's dead wrong here. I have a few bones to pick with the Coveyesque nature of his argument.
Is Covey the Right Benchmark in an Era of Big Data?
First up, I shudder at the notion of the data visualization as a "project." To me, that term connotes that there's a definitive start and end date. Projects or initiatives by definition end at some point. As I research dataviz for my forthcoming book, it's obvious to me already that the most progressive organizations don't fall into this trap. DV is an ongoing and evolving process, a critical tool for managing Big Data. Companies like Netflix continually build new data visualization tools based upon evolving business needs. The company is never "finished."
Project? What project?
Along these lines, beginning with the end in mind is often the last thing that organizations should be doing. Of course, any reporting tool - and I use that term pretty loosely - should produce meaningful business measures. It doesn't hurt to know what you'll want to see a few months after activating it. While no one can predict the future, it's pretty reasonable to expect some level of basic reporting.
Except interactive dataviz tools should be considered reporting tools, at least in the traditional sense. In fact, in an era of Big Data, do we really know what we're looking for all of the time? Interactive dataviz applications allow for exploring and refining of original questions. Curiosity is a good thing. We should embrace it.
Simon Says: Explore, Experiment and Iterate
We live in an era of Big Data. If you know what you're looking for, it's not terribly hard to find it. The more important questions are:
- Did you find the right things?
- How do you know?
- Will those still be the right things in a few months?
Beginning with the end in mind seems to be a relic of a bygone era - one of Small Data, KPIs, simple dashboards and standard reports.
Be curious. Begin without a clue about where you're going. You might wind up in a better place.
Feedback
What say you?
4 Comments
Covey says, begin with the end in mind. Simon says, explore, experiment and iterate.
In theory, great PHILosophy. But most purchasers these days are looking for ROI figures which are still tied to "small data" concepts as you put it.
Any suggestions as to how you would sell big data technology based on "Explore, experiment and iterate?" How are any of my customers going to be able to justify a technology investment by stating internally that they don't have a "clue about where they're going. But they might wind up in a better place."
I just can't seem to visualize that scenario.
Your thoughts?
I like where you're heading here Phil. I'm hoping you might be so kind as to put a more practical "sales' framework around your concept.
Thanks for the comment, James.
Well, if you have "sell" Big Data internally, maybe the time isn't right. Look at what Amazon, Apple, Facebook, Google, Twitter, Netflix, and scores of other companies can do. Big Data represents a big part of their success.
I've never been in senior meetings at these companies, but I sincerely doubt that precise ROI calculations rule the day.
Phil, we are both right. My comments were about traditional BI + OLAP. You are talking about data exploration / discovery, etc. Different use cases.
If you think about begin with the end in mind from the behavioral sense, as the habits are intended, I believe this habit can still apply in the world of big data and data exploration/visualization. The end in this case is NOT traditional in the OLAP/BI sense (a report with set metrics), in fact the end is somewhat uncertain (not sure what I'm going to find out). So beginning with the uncertain end in mind you can change your behavior and processes to support it (I get all the data first because I don't yet know what I need, if IT lets me, as an example).
Just thinking out loud.