The ideal business analytics practitioner is ridiculously multi-skilled. They need to be an expert mathematician. They need to intimately understand data warehousing and how to deal with Big Data. They need to understand not only their own industry sector and business, but also be able to translate innovations from other sectors. They need to be expert communicators and financial modellers. They need to be evangelists. And probably more than anything else, they need to be comfortable persuading and negotiating to deliver real organisational transformation.
Finding someone like that's a tall order. Is it even possible?
There's a concept in economics known as 'signalling' - just like it sounds, it's a way for individuals to send a message about their skills or backgrounds. A peacock's plumage is a prime example; any peacock that can maintain an apparently 'expensive' and full set of feathers clearly has better access to resources than another peacock with a mottled tail. If you're a peahen looking for a good mate, a full set of feathers is a pretty good starting point.
To be effective, these signals need to be cheap for those with the right characteristics to send while being expensive for those without those characteristics to capitalize on. Otherwise, we experience a phenomenon known known as 'free rider syndrome' where everyone takes advantages of those signals, reducing the ability of those searching to be able to differentiate those with the right stuff from those without.
The importance of signalling applies everywhere, even as broadly as pirates. Flying the Jolly Roger communicated with extreme clarity that resistance would be met with no mercy - should the merchants fight back, they'd be slaughtered. Should they surrender peacefully and not fight back, they'd be allowed to go free.
Because of this, merchant ships tended to surrender without mounting a fight when confronted by pirates, saving the pirates from expensive injuries and ship repairs. Given the impact of the Jolly Roger, it's interesting that privateers (and other state-sponsored raiders) were so reluctant to take advantage by also flying the flag. The thing is, flying the Jolly Roger came at a high price - known pirates would be hung if caught by the authorities. As a criminal who would be hung anyway, this threat meant little.
However, for a state-sponsored mercenary interested in reducing their operational costs while raiding merchant ships from other states, temporarily flying the Jolly Roger to inspire fear came at a high cost; if seen and marked as a pirate, they'd lose their contract and likely be killed. It's an interesting historical note that merchants were far more likely to surrender when threatened by a pirate than when treated by a mercenary despite the mercenaries often being better equipped. If you're interested in knowing more about pirate economics, do check out The Invisible Hook: The Hidden Economics of Pirates; it's a fascinating read!
As a profession, business analytics has yet to develop those signals. Knowledge of algorithms is one thing, experience delivering outcomes another. Having specific products listed on one's résumé communicates little with technology platforms becoming so broad an comprehensive. For example, knowing how to use 'SAS' could mean anything from experience with data warehousing, high-performance computing, time series methods, or business intelligence and reporting.
As the profession becomes more formalized, these signals only become more important. While the investment required for a high-performing individual isn't cheap, it pales in comparison to the time and opportunity cost of failure. Getting things wrong because you chose the wrong person can mean lagging your competitors by anywhere up to a few years!
Given the breadth of skills required to succeed in business analytics, these signals are necessarily complex. In practice, they need to communicate technical skills, academic skills, and experiential skills. Unfortunately, when all we have to compare candidates is a résumé that's easily faked using buzzwords, the cost of 'faking it' is relatively cheap, especially when demand's high enough in the market to make it relatively easy to find a new role when things go south.
Recognized signals help everyone. They help those with skills differentiate themselves. They help those hiring those skills protect their investment. And, they help define a growth path by which those interested in entering the profession can develop and qualify their skills.
Sadly, it seems that we have yet to develop these generally recognised signals. So, how do you qualify experience?
Besides providing a quick, simple definition ("big data is all about creating, analyzing, and managing large data sets"), The 3-page report does a nice job distilling what's driving this trend. It also explains where enterprise applications are investing to embrace big data, and what other tools are needed within organizations to embrace big data.
Continue reading "Big data reality check"
If no knowledge is captured, then we miss the opportunity to learn from our mistakes and share vital information with others. There are two key areas we can address that help solve this dilemma – knowledge management and process improvement. To help you get started, I’ve got four books that I would like to recommend:
Continue reading "My summer reading selection: Process improvement anyone?"