The Wavelength: Business Intelligence

0516wavelengthpanelBusiness Intelligence, or BI, is a term we hear everywhere these days. From entertainment to manufacturing, businesses are collecting massive amounts of data and analyzing it in order to make better business decisions and improve customer service. We asked some executives who are very familiar with BI a few questions about what trends they see in the field and how their own businesses apply it.

What new trends can we expect to see in BI this year?
Marc Altshuller: A modern BI platform should allow nontechnical users to automatically execute full-spectrum analytic workflows from virtually any information source – from traditional spreadsheets to multi-structured documents such as PDFs, text reports and web pages. The platform should also enable self-service data preparation, because it is one of the key components in an effective analytics strategy.

However, in many instances an organization’s data is diverse and rarely presents itself in a format that is accessible or needed to perform analytics. Data comes in a wide range of formats, which often require IT intervention to make it usable to the line of business professional or citizen analyst.

Many organizations are spending entirely too much time preparing data and not enough time analyzing it to uncover critical business insights. This is why it’s essential that today’s BI platform enables businesses to quickly access any data source and automatically convert it into structured data for analysis. This will not only expedite time to insight, but it increases the likelihood of uncovering new insights that have the potential to transform the business.

John Doyle: The field of BI is fast-paced with new trends evolving regularly. We expect cloud BI to thrive with hybrid options — data in the cloud is scalable and easy to set up. When adding unlimited access to on-premises data, cloud BI solutions will reach their full potential. Another trend is the democratization of data. Data science and advanced analytics will not be just for data scientists anymore — with [statistical programming language] R and predictive capabilities becoming pervasive through data analysis and visualizations, the realm of machine learning will become accessible to a broader group of regular people who are not trained as data scientists.

Nick Halsey: We see two major trends: Streaming analytics and analytics across multiple data sources. BI to date has been mostly about batch-oriented analysis of data based on what happened in the past. The new world of streaming analytics is all about real time and ad-hoc analysis that leverages a modern technology stack. Real-time data is obviously best handled as a stream. But it’s possible to stream historical data well, just as your DVR can stream Gone with the Wind or last week’s American Idol to your TV. This distinction is important, as we believe that analyzing data as a stream adds huge scalability and flexibility benefits, regardless of if the data is real-time or historical. We are also seeing a huge increase in desire to engage in interactive and intuitive data visualization and exploration with data coming from disparate data sources. Organizations want to fuse this data and analyze from multiple dimensions.

How has the cloud affected BI?
Altshuller: A modern BI platform should allow users to securely connect to corporate data from the cloud. By moving analysis to a self-service cloud-based model, business users can streamline analytics projects without investing in expensive and complex IT infrastructure, and ultimately gain insights on critical data, anytime, anywhere, faster.

Doyle: The rapid growth of BI speaks in part to the explosion of data, which IDC and others estimate could reach 50 zettabytes or more by 2020. Given its massive scale, the cloud serves as the only practical place to glean and disseminate insights from the diverse data surrounding us. In addition, the disruptive nature of SaaS (Software as a Service) is making BI more accessible. Cloud-hosted, business intelligence and analytics services are now permitting business users to directly connect with and gain insight from their business data. The result is that more people can connect with and gain insight from their data, faster and more simply than ever before.

What role does visualization play in BI?
Altshuller: Data visualizations are an important component of a modern BI platform and data driven discovery. Turning data into easy-to-understand graphs and charts so analysis results can be communicated to others is essential. One of the things we were hearing from customers, though, was “Are these visualizations giving me key insights to make the right decision or are they just confirming biases or pre-existing assumptions?” This is where Watson Analytics is different. As an easy-to-use smart data discovery service, it guides users through exploring their data, automates predictive analytics and makes creating dashboards, reports, and infographics almost effortless – without requiring a professional analyst. By statistically interrogating the data, users are assured of unbiased, objective insights. For example, the visualizations and insights generated within the solution can help marketers quickly spot customer buying and behavior patterns and untapped campaign opportunities.

Halsey: The best visualization is the one that allows a normal human with understanding of a business system to quickly see how the visuals match up to the business process. The dirty little secret of the ongoing “data science” boom is that most of what people talk about as being data science isn’t what businesses actually need. Businesses need accurate and actionable information to help them make decisions about how they spend their time and resources. They want to do so in the context of their day to day business activity at any time, using any device, including mobile and wearable. They further want such visualizations to be providing guidance towards next steps or suggesting what others may have found useful in similar contexts. The key difference that we are seeing in this context from the old world BI days is that end users expect visualizations to be smart and insightful – so that they can make decisions faster with better quality information.

Jun Ho Son: Enormous, on two levels — using graphical representations of the data, we can discern obvious trends and the categories of service calls we affect more clearly, and also allow customers to more clearly see this. First, we gather and correlate all the power anomalies we collect on our appliances with service call data provided by customers. We also focus on visualization to enhance usability. For example, the ability to map power data with floorplans of machine-in-field (MIF) deployment makes it far easier for technicians to identify and understand hot spots quicker, even before there are issues. We are trying to make sure that we are enabling usage and insights across multiple levels of an organization.

What are some of the biggest issues affecting BI today?
Altshuller: Data analysis is shifting to a self-service model. By 2018, “smart data discovery,” which includes natural-language query and search, automated advanced analytics and interactive data discovery capabilities, will be the most in-demand BI platform. The smart data discovery user experience will enable mainstream business professionals to get insights such as customer segments, predictive drivers, and outliers from their data.

Cognitive computing is helping people understand, reason and learn from their data in new ways. IBM Watson Analytics helps individuals unlock the value of data they already have in their systems, as well as new valuable external data sources (e.g. Twitter and weather data) they may not even know they need. By bringing as much data as possible to the problem at hand, users can answer their toughest questions and embed insight and expertise into every decision they make.

Russ Gould: There are five issues affecting BI today:
1) The need for governed data discovery. Traditionally, IT owned the data and its distribution to the business. IT was forced to do more with less, which created severe impacts on resources and businesses were unable to get the reporting they needed. Data discovery vendors entered the market, focusing their efforts on providing businesses with the ability to do their own data crunching and analysis. While this freed up IT from the constant demands for reporting, it opened up other concerns, such as the lack of consistency and data governance. Many businesses are now facing a proliferation of BI tools within their organizations, but have little control or management of the data used to make strategic business decisions.

2) Maximizing the effectiveness of available data. Data is everywhere. The challenge is not the velocity, the variety or the volume — i.e., big data. It’s not the amount of bytes. The key challenge for BI users is leveraging the granularity available from an organization’s vast sources and the web. Winning organizations are able to harness the knowledge gained from deep within these rich, vast stores to make better decisions and take action.

3) Analytics and BI adoption throughout the organization.
We’ve come a long way. However, after 10 years of BI being the top technology investment priority, CIOs and CEOs still project analytics as a top three priority for the next five years. Organizations have a long way to go too; most are in their infancy along the maturity curve adopting analytics and business intelligence. Analytics is now everyone’s business, not just the data scientists, but making analytics pervasive in their organization is a challenge. Leading companies use analytics built into the fabric of everyday operations to create business value. Analytics will be pervasive in market winning organizations.

4) Enough people with the right skills. There is already a shortage of data experts and data scientists to support the millions of business users. This growing disparity is getting worse. There will always be a need for the data scientists, but there will never be enough of them. Enterprises need to find people within their organization with an aptitude for analytics, data literacy, and some data science skills. They are going to have to develop employees. Just like organizations train new employees on ethics and corporate policy, they need to train employees in analytics. Think of it this way, just because someone has desktop publishing software and creates a flyer doesn’t mean they produced something of value or that they are now a communications expert. The same holds true for analytics — an employee can create dashboards, it doesn’t mean they produced something that’s valuable toward achieving department and company objectives.

5) IT not trying to standardize on one BI solution. It is a BI market reality that there are many choices. The ability for organizations to decide on which BI tool meets their needs is becoming increasingly complicated with technology pressures. There is even more pressure from businesses for agility and speed, greater pressure to drive competitiveness in all levels of the organization, limited skilled resources in the market to meet demand, and a plethora of choices that make sizing up the market even more challenging. CIOs face huge challenges in balancing the needs of their internal customers, as well as ensuring the business can meet the needs of external customers.

Son: Two of the biggest issues affecting BI today are getting customers to understand how use of BI can immediately and directly impact their business and data security. While most businesses conceptually understand that the ability to utilize more data can be beneficial to their businesses, they often don’t know where to start. BI doesn’t have to be highly complicated, super advanced analytics that you see in ads — this tends to intimidate people from getting started. Simple insights can also yield more immediate, tangible benefits. For example, power profile data gives technicians the ability to more quickly diagnose whether power is or is not a potential root cause of failure. This data can then be aggregated at the cloud for further analysis across [installations] and for predictive maintenance. The point is that our customers can get started right away with something their techs can immediately use and derive value from.

Data security is the other big inhibitor. While our solutions never handle customer data, we need to constantly show that the protections we build into our solutions don’t unwittingly create a backdoor issue for them. We spend a lot of time working with customers, not just educating them on our technology-based protections, but also on working with their existing security processes and protocols to ensure monitored access and usage rights. This is a never-ending quest to not only try to build in the best protections but also the right remediation procedures and stopgap measures.

How do you provide access to BI for your clients?
Altshuller: The rapid ascent of Watson Analytics into one of the most popular self-service analytics platforms in the world has been fueled by its unique ability to put cognitive capabilities into the hands of business users and enable a new era of unbiased analysis.

For example, Kristalytics, a marketing analytics company in Texas, is identifying trends in consumer behavior and building customized reports for its clients using Watson Analytics prediction and data visualization capabilities. By integrating and analyzing a mix of client revenue data with publicly available consumer data, the team provides deep insights on market behavior that its clients can use to inform their marketing strategies. The company recently moved its data into Watson Analytics to help build predictions and gain new insights that it could share with clients, and immediately identified reasons behind one client’s underperformance: Watson Analytics drew correlations between gender, age, and proprietary Kristalytics data that showed that the client was underperforming with males. By proposing lines of questions to follow next – and guiding the analysts to look at gender age breakdown – Watson Analytics helped reveal insights in a few days, rather than months. Watson Analytics visualizations are also helping them present their data to clients in new and compelling ways by building customized infographics for each client.

Doyle: Microsoft’s approach is to make it easier for our customers to work with data of any type and size — using the tools, languages and frameworks they want to — in a trusted cloud or hybrid environment. Microsoft offers a comprehensive set of data services for customers — including Power BI, a cloud-based business analytics service that enables anyone to visualize and analyze data with greater speed, efficiency, and understanding. It connects users to a broad range of live data through easy-to-use dashboards, provides interactive reports, and delivers compelling visualizations that bring data to life.

Halsey: Customers can use Zoomdata on premises, in a private cloud, on any of the major public cloud platforms or in a hybrid deployment. Zoomdata’s patented Data Sharpening technology offers the industry’s fastest visual analytics for real-time streaming and historical data, while Zoomdata Fusion enables the joining of big data with other modern and traditional data stores. Zoomdata is the first business intelligence solution natively architected for cloud and on premises deployments using an optimized microservices delivery approach that delivers visual analysis of big data sets in seconds. Customers can also sign up for free trials either using Docker or Ubuntu for on-premise deployments as well as for cloud deployments using Amazon AWS, Google Cloud Platform, or Microsoft Azure.
Son: Our primary access to BI for our customers is in the cloud, with secure, domain-structured access controls that enable us to tailor the data to how customers view and run their business. APIs are available to port power profile data to existing dashboards that customers may already use. Real time notifications of power anomalies can be delivered via email. Cloud data is accessible via any internet-enabled device. For ease of use in the field, we have apps and optical readers for Android and iOS devices. For customers that do not allow usage of clouds due to regulatory issues, we also can instantiate a private cloud behind customers’ firewalls.

Talk a little bit about IoT data and BI
Son: We provide immediately valuable protection and usable data at the point of power usage and provide an immediate ROI vs. a standalone sensor. The challenge for sensor-only solutions is justifying their ROI within a reasonable payback period, kind of like when you’re trying to decide between buying an expensive LED bulb vs. a standard incandescent bulb. The LED may save you money over 10 years, but it costs at least five times as much. Beyond what we’ve described for field technicians, there are benefits at the fleet and predictive levels. We’re working on integrating new types of sensor data at the edge, like humidity and temperature, which customers believe are important to their operations. We want to work with the ecosystem of technologies and sensors, as well as developing our own proprietary technology. Service data can also be combined with sensor data to provide strong predictive analytics. With these systems, businesses can expect increased revenues, differentiated services and better customer experience.

This article originally appeared in the May 2016 issue of Workflow.