Being a data-driven organization requires a combination of analysis and predictions. As we look at the last several years, big data has been the dominant force, moving from a consumer-based requirement to a must-have strategy for enterprises to stay competitive. This data revolution is seeing accelerated technology adoption curves and is forcing job roles to change, strategic investments in technologies like Hadoop and data quality/governance, threats to incumbent BI vendors, just to name a few. 2015 will see big changes in enterprise information management, as organizations transform every employee into being data-driven, as analytic use-cases become more complex, and as vendors try to predict the next need of the market.
Prediction 1: The lines will blur between data scientists and data analysts
Harvard Business Review declared data scientist as the “sexiest” job of the 21st century. But the role of data scientist is evolving rapidly from predictive analysis and code writing to someone who is able to to communicate in language that all their stakeholders understand. Likewise, the world of the data analyst who prepares graphs and visualizations from various data sets is also being disrupted.
As business requirements for data analytics and business intelligence tools emerge, the distinctions between data scientist and data analyst has become fuzzier. In 2015, I expect these job responsibilities starting to merge into a new class of data expert who possess the reporting and visualization skill sets but is also capable of understanding “the complexities of the new NoSQL databases like Hadoop and can marry data from many sources and types together to produce useful and profitable predictive models,” as Gartner articulated well.
Much of this has been driven by the democratization of business intelligence tools like Tableau, which has made it possible to transform complex data from back-office applications into simple-to-understand data visualizations.
Microsoft will likely achieve significant economies of scale in its cloud BI offerings, based on the Power BI suite for Office 366, which will result in considerable pricing pressures. Another potentially disruptive force is Salesforce Wave, its cloud analytics platform. If it lives up to its potential, Wave could wash out almost half of the BI incumbents. But Salesforce Wave will not fully crest unless they take a modern approach to how they help their customers solve the integration and quality challenges.
Prediction No. 3: Data preparation replaces Big Data as the hottest topic in BI
2014 was a formative year for data preparation, with a broad cross-section of contributors helping to define it as a distinct category within ETL. The data preparation market itself has been enabled and advanced by major technological breakthroughs in areas such as semantic algorithms, distributed computing, machine learning, and cloud computing.
The coming year will also see the industry delivering data-preparation technologies at scale for all users. The proliferation of ungoverned data will become so large and problematic that companies will need to shift their emphasis on data standardization, governance and data preparation.
The emergence of an all-in-one data prep stack will arise as CIOs and CDOs demand new approaches to end-to-end information management. The industry will respond by moving from delivering fragmented, point products to cohesive platforms built on elastic architectures that can scale and delivered as cloud-based and on-premise.
If the entirety of the enterprise IT landscape were comprised only of Web 2.0 companies like Google GOOGL +2.25%, Yahoo YHOO +0.00%!, Facebook, and Twitter, Hadoop would be considered an unqualified success for Big Data projects.
Hadoop faces a tougher time however in demonstrating its applicability and value inside Big Data projects across the larger enterprise market, which constitutes a much larger commercial opportunity. The enterprise software market alone is estimated at nearly a $100 billion per year, growing at almost 5%. The challenges Hadoop faces in the enterprise market are varied:
- Many Hadoop projects inside of mainstream enterprises are still largely “science projects” that have not been integrated into production environments;
- The major Hadoop distributions (Cloudera, MapR, Hortonworks, Pivotal, and IBM etc.) are still largely not interoperable with each other;
- As Hadoop encroaches further into mainstream enterprises, incumbent database vendors are responding more aggressively in protecting their primary turf.
Some other major Hadoop developments I foresee in 2015 include MapReduce being reduced to the ash heaps of history while Hadoop Spark having a banner year and becoming a mainstream technology.
Prediction No. 5: Marketing will be the primary driver of BI decisions
Traditionally, the IT department called all the shots when it came to all technology decisions. Then disruption happened in the forms of cloud computing and smart mobile devices, threatening IT’s hegemonic influence over the lines of business (LOBs). Self-service technologies have given rise to “shadow IT” and the shifting of the balance of power away from IT to the LOBs. According to Gartner, more than half of all budget spent on BI and analytics today are driven by the LOBs.
The power organization to watch in the BI and analytics space in 2015 is Marketing, as it attempts to reposition itself from a cost center to the growth engine of the company. BI and analytics are enabling platforms for data-driven marketing, along with a dizzying array of other tools and services that make up the marketing technology landscape today.
IT organizations who do not cooperate will be subject to whiplash as Marketing zips around them to get what they need elsewhere.
Prediction 6: Internet of Things Gets Real for B2B
One of the hottest “inventions” debuted at the 2014 Consumer Electronic Show (CES) was IoT toothbrush, enabling parents to see how long their children brushed. The number and variety of devices that will comprise the IoT is staggering and growing, everything from the Fitbit and Apple Watch to even the most economical car.
For instance, in the construction industry, workers can now get supplies brought to them instead needlessly going up/down ladders and risking injury. Imagine the insurance company who would offer discounts on premiums to the construction firm using an IoT ladder over an older model. Not so different than a utility offering discounts for purchasing “smart” appliances. An insurance company could correlate the data coming from the ladder against claim data and against risk data. Suddenly, the IoT data is ONLY contextual, and therefore valuable, if it can be viewed against other meaningful data coming from data warehouses.
Source: GUEST POST WRITTEN BY Prakash Nanduri, Cofounder and CEO of Paxata. DEC 19, 2014 @ 09:15 AM