The Data Economy: Does the growing influence of CMOs portend a wave of Big Data apps?

uncertain road path street question mark predictive analytics direction futureThe Data Economy is an occasional analysis column by Wikibon Senior Analyst Jeff Kelly covering the business of Big Data.

In Wikibon’s recent market forecast, we determined the current Big Data applications market stands at $1.7 billion. That sounds big, but the reality is that most of that revenue is derived from the sale of data visualization and business intelligence tools that are used to present Big Data, as well as analytics software for doing the underlying analysis.

But as for net-new Big Data applications – that is, discrete, packaged applications for business users that leverage Big Data analytics under the covers to solve specific business problems and challenges – there are very few currently even on the market, let alone generating significant revenue.

That’s unfortunate because I fundamentally believe the real value of Big Data will only be unlocked through these types of applications. While the underlying infrastructure and database layer – including Hadoop, NoSQL data stores and MPP analytic databases – are critical enablers to Big Data success, they are just that: enablers.

What’s holding us back?

 .

The reason I believe we have yet to see significant traction in the Big Data application space is that building Big Data apps is just really hard. Sounds simplistic, I know, but there’s no getting around it. Typical enterprise applications are deployed on relational databases and generally deal with a single, known data source. Big Data applications are developed on more complex, distributed infrastructure, integrate and analyze multiple data sources, and must present insights in ways that make sense to business users so they can take relevant actions.

And that’s just the technology component. Big Data application development also requires Data Scientists – the ones doing the hard analytics – to collaborate with application developers – the ones tasked with productionizing insights Data Scientists discover. And don’t forget the domain knowledge required to tackle the hard problems Big Data apps are trying to solve, be it detecting fraudulent financial transactions or determining next-best-action for customer interaction.

Expectations for market maturation

 .

I’m optimistic, however, that we will see new Big Data applications coming to market over the coming months and years. For one, the underlying infrastructure, such as Hadoop, is maturing rapidly and being hardened for the enterprise. Another leading indicator, I think, is the growing influence and power of CMOs and marketing departments in making technology purchasing decisions. Marketers aren’t interested infrastructure, but they will pay for applications that solve business problems. And smart vendors go where the money is.

Case in point is Teradata. The company is known for its data warehousing appliances, but the company is also moving up the Big Data stack to the application space, specifically focusing on marketers. Its Real-Time Interaction Manager application, for example, is able to blend and analyze multiple real-time and historical data sources to drive targeted marketing campaigns to customers.

I’m not ready to declare 2014 the year of Big Data applications (I tried that once before in 2012, and we know how that worked out), but I am hopeful that new apps are on the way soon.

I chatted with my colleague Stu Miniman about this topic in a CUBE Conversation this morning. Check out the full conversation below.

photo: milos milosevic via photopin cc