Manage big data properly, don't do it 'just because', warns Pentaho
Big data skills can be learned, but good implementation strategy is required for an analytics strategy to work, argues Pentaho's Wael Elrifai
Poor management processes are holding back business which are attempting to successfully implement a big data and analytics strategy, a big data and machine learning expert has warned.
Wael Elrifai, EMEA director for enterprise solutions at Pentaho, made the comments in an interview with Computing ahead of our sister-site V3's Big Data Summit, during which he'll be advising delegates on key big data trends.
Elrifai argued that it isn't a lack of skills that holds back a big data implementation because "programming is programming; if someone learns on C++, they can learn other languages, that's not an issue", but rather there are problems around understanding the process side.
"When you do an implementation, you maybe have one piece right but not another piece, or have this technical bit right, but you don't have the correct process in place because you don't have a data scientist or management buy-in to implement changes," he explained.
Essentially, despite some of the hype surrounding big data, the operations side of it still needs managing - a business can't just install a system and hope for the best, he said.
"Because when you're doing predictive methods - or even prescriptive analytics - when you're getting feedback about support systems, you still need to have a management process to implement them," said Elrifai.
"If you built a real-time predictive system and if you don't have a management process to implement the recommendations, well then that's not going to get you much ulitity, is it?," he continued. "Part of the problem is that companies have been lacking knowledge about how to put these projects together".
Elrifai also told Computing that a "me-too culture" around big data - wanting to use it just because it seems like everyone else is - results in businesses trying to use analytics, but for ill-defined ends.
"I think another part has been a ‘me-too culture' of just implementing things because they're there without having a clear understanding of what your use-case is," he said.
However, Elrifai argued that it's obvious why some firms act in this way; because they want to get the jump on their rivals when the time is finally right for them to harness big data and analytics.
"Because companies do need to build their internal skillsets, even if you don't have a use case for something yet, sometimes it's justifiable that you want to get ahead of the problem to make sure you have the skillsets when your competitors start," he said.
The V3 and Pentaho ‘Big data for big business' webinar takes place at 11am on 3rd November 2015.