At the recent “Big Data, Analytics & Applied Machine Learning” 2015 Israeli Innovation Conference hosted by IGTcloud, Viola Ventures’ General Partner Ronen Nir raised “5 Big Questions about Big Data” (see the video of his presentation below).
Excerpt from the presentation’s intro:
“Big Data has been a big focus for us at Viola Ventures (formerly Carmel Ventures) for the past few years. We spend much of our time meeting companies, talking to customers and sitting in board meetings in an effort to understand what our companies’ customers’ can gain, so we have a good understanding of the end user’s perspective of the value of big data. While this is clearly one of the mega-trends in the industry, the ecosystem raises some very significant questions that I thought would be good to share.
These questions aren’t meant to be provocative but I think that as an industry we should all be trying to answer them. Honestly, I’m not sure I have all the answers so I may be wrong in some of my assumptions, but based on what I’ve heard from other industry experts I know that I’m not the only one who has these questions.”
The five questions about Big Data that Ronen talks about in the video are:
1. Is it working? Not so much in the technical sense, but rather in terms of providing the perceived value. In his talk Ronen analyzes some of the vertical implementations of Big Data such as DNA sequencing payment fraud, advertising, cyber security and IT health monitoring.
“The issue of false positives is becoming one of the biggest issues of driving business value out of big data application.”
2. Hadoop: Is it the disruptor or disruptee? Hadoop is fast becoming the de facto standard for offline BI, yet the initial architecture is still far from providing a complete solution for challenges such as complicated installation and support, high fragmentation in query tools, and SQL queries. While still far from mature, we’re already seeing technologies and architecture that aim to completely disrupt the Hadoop ecosystem.
3. Is open source shrinking the market? Many big data solutions rely on open source, but is open source good or bad for the market? Open source may not diminish the overall market size but it’s highly likely that it will significantly change it. Since the core is free, much of the revenue that used to come from selling software licenses is now coming from professional services, support, training and other areas. But is this necessarily a good thing? It’s debatable.
4. Is Big Data the cloud anti-trend? Apart from the obvious security and regulatory problems, there were other hesitations about moving big data to the cloud, including the high cost of data transfer and the changes in architecture. But these two issues are starting to change: Migration tools are more mature (AWS/VPC, IBM/Softlayer and other cloud enablement tools), Architecture is becoming an extension of the enterprise, and native Big Data Cloud Services are emerging.
5. Will services inherit software licensing? Two major problems facing large organizations that deal with Big Data include the fragmentation of vendors and the scarcity of Data Scientists. The solution will be to completely outsource many of the disciplines, for example IT management and cyber security.
“Big data today is a services business and it’s going to continue being a services business. Even in 2020 when it will be a $50 billion industry, more than half of it will still be a services business, which will most likely move slowly and require a lot of professional services and implementation, and I believe that we must change this. I think that for the benefit of all of us, we must turn big data into something that has business value behind it, which we can deploy faster, and that is really a software and technology enabled solution.”