Do the uncharted seas of Business Intelligence lead to new lands or are we sailing off a cliff?
Let’s send a 21st century corporate IT director back in time. I’ve chosen late July 1492. The setting is a harbor-side bar in Palos de la Frontera, Spain. Our intrepid IT director is sipping sangria with Christopher Columbus.
At the end of the pier you can see dockworkers loading supplies onto Columbus’ ships, the Pinta, Niña, and Santa Maria. Let’s listen in.
Columbus stares out across the water. “I’m going to find a westward path to India.”
“Tired of working in the cheese shop, eh?” says the IT guy. “I don’t blame you. Will you need a navigational system upgrade for the voyage?”
“I’m not sure what I’ll need. This is a journey of exploration. The tools we have today will likely prove inadequate, but in ways we can’t predict sitting here in port on a sunny day.”
“I hear your brother makes maps. Need any maps?”
“Maps are for followers.”
“Hmmmmm.” IT guy sips wine, admires label, wonders whose budget this is going to hit.
“Well, we can spec out a project plan to make sure your ships are seaworthy.”
“I’m sure you could, but at the end of your project my ships would be seaworthy only in the sense that we understand seaworthiness today. That might not be good enough.”
“Maybe you should take a priest on this trip instead of an IT guy.”
“Now you’re beginning to understand.”
In that snippet of dialogue we see the stark difference between exploring the new and perfecting the known.
I first ran across this idea while preparing a blog post on the pundits’ 2013 predictions for the BI space. A few industry watchers talked about how Big Data could usher in a golden age of exploration and discovery if we let it.
At the time I thought this was an interesting notion, so I tucked it away in the “worth revisiting” file. In the meantime, the venerable Harvard Business Review (HBR) weighed in on the topic in a piece called Why IT Fumbles Analytics. The authors came at the subject from a slightly different angle, but it’s a good read, albeit not a quick one.
Here are the salient points. The conventional approach to IT projects works well given the right parameters. For example:
- Relevant data has been identified
- Data is understandable
- Data is controllable, structured
- Workflow is repeatable, process-oriented
- Timelines and milestones are bounded, definitive
In short, when you look at a project and can confidently proclaim that it amounts to automating a series of well-defined manual activities that humans do poorly and computers do expertly – well then, let’s just say that’s when the traditional IT approach is going to work beautifully. Think of it as taking unruly data out of the world, sedating it, and bringing it back into the light on a leash. All well and good.
However, BI projects are different animals in some respects. They succeed by unchaining human intelligence, not by eliminating it. The challenge is to improve the way decisions are made.
Now let’s compound the problem. The world is messy, and Big Data gives us a vast new ocean to play in. There are no rules. Don’t miss the irony here! Technologies that were supposed to help collect and control data are now causing a data deluge.
Conclusion: the conventional IT project design philosophy doesn’t work in reverse. Getting data, especially Big Data, out of the cold machine and into warm human hands so that it can be examined, debated, discarded or turned into usable information is unconventional. Wild, hungry data is being invited into camp.
HBR says this:
IT projects don’t usually encourage people to look for new ways to solve old problems.” Moreover, they don’t usually encourage people to look for new ways to identify and exploit new opportunities.
The authors go on to suggest ways to deal with the problem:
- When the goal is to support the discovery of new solutions and new opportunities, build your team with an eye toward more “I” and less “T”
- Frame your project like a clinical trial, not scientific research
- Make sure everyone knows it’s OK to use words like hunch and gut
- Focus on insights that might help the business and don’t get hung up on data pedigree. As Ray Kroc said, a good idea doesn’t care where it comes from.
Bon voyage. Remember to load up on fresh water and limes.