I have noticed that little has been written about the set-up needed to succeed in Big Data projects.
To increase chances of successful implementation it is important to have the players well connected: Business, IT and Data Science.
Five steps to connect these three teams:
First it is key that Business, IT and Data Science teams understand the basics of each other’s language. That is not a trivial challenge, considering that it is even hard for specialists to keep up with the lingo and developments in their own domain.
Secondly, instead of rushing towards solutions and road maps, sufficient time needs to be spend on : ‘What is the real problem to solve?‘.
Thirdly, use-what-you-have. Look at the tool-set and capabilities you have in-house. In larger companies especially there are quite often sufficient tools available to make a first step. Look for the analytical brains in your teams. They might not have Data Science on their business cards, but they could have the potential.
Fourthly, go for the low-hanging fruit. For big data projects to be successful, it is essential to make many iterations. Find an area that jumps out, focus, keep scope limited, and start trying out.
Last but not least, mix and mingle. Don’t make this a project for one group, make it a joint initiative from day one. Try to have the team in one room to maximize learning, feedback and communication.