Christian Prokopp, one of the Principals of Big Data Partnership, took time out from helping our clients along their Big Data journeys recently, to write up a helpful piece here about some of the factors that derail Big Data initiatives. And some guidelines for avoiding those problems and for fixing them if you’ve already met them.
An observation quickly picked up by commenters on Christian’s blog is that key to success in exploiting Big Data is a good understanding of what you’re trying to achieve. As commenters pointed out, you’ll do better at most things if you do a little planning. But i think this is doubly important with Big Data, because with the advent of Big Data technologies the solution space has changed; it now offer new options for all kinds of use cases; offloading slow-running analytics from data warehouses and replacing ETL architectures being two common ones.
But more importantly, and with more crucial impact on businesses, the problem space has changed. Use cases that were not addressed at all can now be served, sometimes surprisingly easily. So for example handling very high arrival rate data and analysing it to produce results in real time was prohibitively expensive. Now it can be delivered cheaply on open source software and commodity hardware. With the rise of social data already on us, how many marketing organizations have a real-time picture of the effect of their campaigns and product launches so they can adjust in flight? A feedback loop that took weeks to tell a CMO if a new product features was being well received can be a feedback loop in minutes.
So it’s not just a matter of good project planning. It’s vital that the business takes a step back and thinks the unthinkable. And does that with technologists who understand what can now be achieved.
You don’t need your hiking boots for the big data journey, but you do need to know where you want to go to, and a map and experienced guide helps.
Comments are closed.