Prakash Nanduri is Cofounder and CEO of Paxata

2016 marks the third year that I’ve penned my business intelligence market predictions for Forbes. Upon reviewing what I wrote last December for 2015, I realized that while many of the predictions had already come to fruition, some are proving too soon to tell. So this year, I’ve decided to focus my predictions specifically around a common theme to better be able to measure myself next year. Overall, 2016 will be “The Year of Information on Demand.”

Prediction 1: The big bang you hear is the explosion of innovation.
The combination of scale out + hyper-converged + elastic cloud architecture + machine learning is creating massive disruption– breaking all existing paradigms for computing – particularly the performance-for-cost ratios. SQL on Spark, for example, allows us achieve 100x the data processing performance at one-tenth the cost of traditional Business Intelligence (BI) architectures, removing the need for relational data warehouses and other old world tools. This new world will create new leaders…

We’ll see the Dinosaurs trying to reinvent themselves by cross-breeding (e.g., SAP + HP, MSFT + SFDC), while some public companies will seek shelter from Wall Street by going private. Teradata, as an example, might be the first to go, and I expect to see others the caliber of EMC + Dell or SFDC + ?!?)

Organizations are consolidating traditional fragmented data integration, data quality, ETL tools into a more connected information layer designed for self-service access and enterprise-grade security and scalability. In order to remain relevant, vendors will attempt to “retrofit” their legacy IT-centric offerings but will be thwarted by the head winds of self-service despite their own forays into providing business-consumable front-end capabilities. The winning designs will be business-analyst centric and built from the ground up for today’s velocity, veracity and variety of data, with transparent governance and multiple deployment models.

Prediction 2: We must adapt or die.
Data preparation will be a critical capability of subject matter experts who, traditionally, relied on others to get data ready for them. In order to transform data into information on demand, people doing risk analysis, customer targeting, security monitoring, marketing or sales operations will need the necessary skills and tools to handle self-service data preparation at scale.

Those who don’t adapt to the modern paradigm will experience big data blunders, including misclassification of data and embarrassing data quality errors, as the gap widens between all of the data and the people who understand how it should be used. Expect to see more stories like these in 2016.

Prediction 3: Whose BI will you buy?
Several trends will emerge in the Business Intelligence market. Spending has come to a screeching halt in traditional BI platform investments. Gartner sees a decline of more than 20% in net new license for traditional operational BI vendors as organizations are now increasing their spend on self-service BI and data discovery tools, such as those from Tableau Software, which saw incredible growth of 77.7%. This paves the way for more disruption:

• The 800-pound data visualization gorillas will subsume basic data blending, relegating that capability as “table stakes” which will leave point solution vendors scrambling to prove differentiation.
• Seattle is the city to watch as MSFT Azure and AWS make big moves in the cloud BI market place. How will Mountain View respond? With Diane Greene at the helm, Google has an opportunity to make this a very interesting competition.
• AWS QuickSight will also disrupt the playing field for first generation cloud BI vendors who are not highly differentiated and can’t compete with AWS margins.

In general, we will see a major movement to adopt Cloud in the financial service industry. In fact, one of the top five financial institutions will announce a Cloud-first or Cloud-only IT philosophy –adopting one or more of the big three players. If the CIA is doing it, why can’t the banks?

Prediction 4: Enough about data – where’s the information?
The 2016 Presidential elections will be the first time in our history where a new approach to information on demand will trump traditional polling as the means for predicting voter behavior and intentions. While micro-segmentation and statistical modeling was done in the 2012 Presidential elections, it required that analyses be done in sample sets in order to deal with all of the variables and data sources. We now have the combination of machine learning, artificial intelligence and compute power to instantaneously build Nate Silver-like predictions without sampling. Picture doing predictions on the entire voting population, bringing together social media data, historical preferences, psychographic, sociographic and firmographic data into information that tells a far richer picture. I vote for that!

While there has been a lot of buzz around the Internet of Things (IoT), people will start to recognize that sensor data is abundant, very dirty and mostly useless. Sensor data, by its very nature, is cryptic and often made unusable after transmission. The more sensor data being collected, the more chaos will be created. It will be critical that we have a pragmatic discussion about how data streams can be cleaned in real time, and why it’s necessary to merge IoT data with static data to create contextual information that is relevant and valuable. Smart City programs are early examples of how the combination of IoT data from traffic sensors, social networks, transportation logs, census and emergency personnel data can drive rapid insights and community response plans to national security events or natural disasters.

Prediction 5: Death by desktop.
Data silos on desktops have created security risks and scalability challenges for years. Because there were no other options, business teams made a conscious decision to use distributed desktop BI tools leaving their IT team scurrying to add controls and security layers around their users. In 2016, we will see success of solutions where security and scalability get “designed in” from the start. This is critical if we are to achieve governed data discovery, a term used by Rita Sallam, research vice president at Gartner.

The shared anxiety around data risk will unify business and IT teams as the new crop of solutions enable simpler administration, transparent governance and agile collaboration.

The movement will be further accelerated as more enterprises are seeing elastic cloud architectures as the mainstay of big IT-standardization decisions. In fact, this will become the litmus test for Chief architects and CTOs: if the solution doesn’t have a built-in elastic architecture, it won’t be deployed enterprise-wide.

Prediction 6: We’ll see just as many CDO’s hired as fired!
The CDO role evolved as organizations wanted to have one throat to choke and one hand to shake for all things data. And while we saw initial excitement in 2014 and 2015 (evidenced by the number of CDO events that cropped up all over the world), the role will have to be better defined with clear responsibility and authority if it is to succeed. This Forbes interview with Jennifer Belissent, principal analyst, and Gene Leganza vice president and principal analyst at Forrester Research, provides a powerful statement: “the democratization of data – both the responsibility for it and the insights from it – across the organization is the ultimate objective.” Look for successful CDOs to focus exclusively on quantifiable business objectives.