Business Intelligence, or BI, is a common term in the business world today. Businesses are collecting more data than ever from a wider variety of sources 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?
Joseph Bradley: Speech as the user interface for enterprise applications. With the growth of conversational artificial intelligence platforms (Siri, Alexa, Google Assistant), we are getting closer and closer to human-level performance in speech recognition. The collaboration between man and machine will evolve. Imagine a desk mounted, voice-activated bot — a BI assistant that provides all actionable information needed to help corporate executives, business managers and key decision makers make informed decisions in real time.
Alex Burton: Enterprises increasingly recognize the potential of BI to help them optimize their workflows, but they’re often unsure how to go about it. We’re seeing leading software vendors respond and differentiate by combining industry-specific expertise and process intelligence in digital workflow solutions. But such capabilities are not universal. It’s important that enterprises carefully assess their options to make sure they select a BI solution that will really meet their workflow needs.
Kevin Kern: Business intelligence has evolved from its origins in benchmarking and business performance analytics to incorporate external data and predictive technologies. This offers a more comprehensive tool for optimizing go-to-market strategies and productivity. As we look ahead, I think we will see the continued evolution of BI as the fusion of traditional BI quantitative capabilities with the structured and unstructured data found in ECM and AI to give a multidimensional view of performance. This expands well beyond just business performance into areas like patient outcomes in healthcare and better modeling of customer behavior in retail environments.
Dan Kogan: The next generation of business intelligence tools will really be about the user experience, which includes the user interface itself. Instead of being an ancillary tool that is occasionally used, organizations are starting to see an age when data is interactive enough that it can become the backbone of a conversation. Now that people have speed-of-thought analytical tools, they can quickly analyze data, mash it up with other data and redesign it to create a new perspective. And as a result of these data conversations, organizations will get more insight from their data
Bruce Orcutt: I am really excited about the accessibility of information and data to the common user. Self-service and real-time trends will continue to make BI available to the masses. This means that staff, teams, and users will be making critical business and process decisions based upon contextual information at their fingertips.
Give us one good example you’ve seen of BI in action
Orcutt: We are seeing new businesses and business models emerging based upon data previously locked in business documents. For example, invoice processing is a big market for data capture vendors; however, we are seeing new businesses emerge offering real-time insights into spend analysis and tax exposure/risk based upon core data locked within invoices. These are not account payable solutions, but rather insight and understanding solutions that mitigate risk, understand costs, and create new revenue opportunities. Startups and established companies are finding new revenue opportunities based upon business data within their documents and customer engagements.
How has the cloud affected BI?
Bradley: The benefits of cloud are hard to deny — scalability, flexibility and reduced cost among many others. Enterprises large and small will continue to use cloud to enable their business analytics and data management deployments. If you think about the need for decision support systems to “learn, analyze, decide and act,” the cloud has performed all these functions to date. With the proliferation of IoT devices, there will be higher demand to move decision making closer to the customer and reducing latency will become a source of competitive advantage. Machine learning and analytics will reside in these devices where they will analyze, decide and take appropriate action. It takes a tremendous amount of data to teach machines the appropriate action; the cloud will act as a learning hub. As an example, an edge device will analyze the temperature to decide if it is higher than 80 degrees. If it determines that the device is higher than 80 degrees, then the edge device will shut down. But the question is, how does the device know that 80 degrees is the shutdown threshold? It is because of the enormous amount of data that has been processed in the cloud through machine learning algorithms that have “learned” 80 degrees is the right number.
Burton: The cloud lowers barriers to entry and ensures that customers of all sizes can capitalize on the power of BI. In this way, the cloud is having a similar impact in BI as it has had in areas such as search and e-commerce. Cloud platforms constantly grow in value because vendors can easily add and update functionality without requiring the user to do anything. The cloud also makes it possible for all users to benefit from the lessons of everyone using the system — for example, by capturing industry best practices and benchmarks.
Kern: The cloud has had a profound impact on BI. First and foremost, it has provided sophisticated enterprise-class BI cost efficiency to almost any sized business. Secondly, very powerful BI tools are able to be used more collaboratively across the enterprise or organizations. Third, it has accelerated the innovation in BI. The roots of BI were in the area of reporting and business analytics, which often were tied into the ERP system the user was running, which meant updates and upgrades were tied to on-premise IT upgrade cycles. The cloud has liberated BI to be able to look at not only ERP but multitudes of data points from other resources to provide better insight into the business. In our own company, we use tools from our ERP provider and higher-level, cloud-based BI tools.
Kogan: Cloud computing has dramatically changed how organizations think about their business intelligence initiatives. Many organizations now live in a hybrid world, featuring data in both on-premise and cloud-based environments. To stay efficient within a hybrid environment, businesses need solutions that work in both situations – especially their business intelligence solutions. BI tools now need to be able to analyze data no matter where it resides, or where it originated from, and tie all of that together. Being able to complete that analysis in a cloud-based environment is even more helpful, giving teams the ability to share results and analyses with their coworkers across the globe in an instant.
Orcutt: Cloud has the potential to be a wonderful delivery mechanism because of the elasticity and on-demand nature. More importantly, as companies become more comfortable with customer data interacting with cloud services and platforms, the opportunity to add specialized processing for context and understanding about the data is more readily available. The cloud provides the opportunity for shared knowledge and concepts that can be applied to discrete data sets allowing completely new business models to emerge around specific data types, content and relationships.
What role does visualization play in BI?
Burton: Visualization technologies democratize your business data by making insights available to decision makers, from department leaders to the C-suite. No longer do employees need to immerse themselves in tables of facts, measures, etc., or consult BI gurus for answers. With visualization tools, anyone can visually see and quickly grasp the meaning in their data and dig in where they see something that is working well or needs improvement. It’s also important to note that visualization capabilities are constantly evolving and enabling more people to derive more insight from business data more easily.
Kogan: Visualization is a core competency of modern BI and analytics platforms. Business intelligence is no longer the domain of dedicated data scientists who can easily digest a large dataset or string of numbers. Business users in marketing, sales, and operations need to easily see and understand their data in quick digestible charts and graphs that fit the data. The real power comes though with visual analytics, where you can truly explore data and ask “what if” questions in a purely visual interface that empowers business users who might not have SQL or scripting skills.
Orcutt: Seeing is believing … visualization is key. If you believe the power will be pushed to the ultimate users themselves, then you have to believe the interaction with the information is critical. Buying decisions are often made based upon the visualization tool and interactivity versus the actual accuracy and effectiveness of the data.
What are some of the biggest issues affecting BI today?
Burton: The biggest issues affecting BI today are data quality, understanding which data will be most useful to analyze, and making data insights actionable. BI capabilities have become so powerful and easy to use that it’s not uncommon for business users to get ahead of themselves. Enterprises need to apply (and vendors need to provide) checks and balances to be sure the right questions are asked, answered, and analyzed, and that the insights are presented in a manner that makes it clear what areas require action.
Orcutt: A challenge that we are working on for customers is the understanding of their data and converting their business documents to business data. Helping companies realize the context of their data and relationships of entities is a very challenging technical problem. BI has done a great job representing structured data and delivering tools to interact and work with data elements; however, we are still working diligently to get the models right, delivering on the benefits of machine learning, natural language processing and AI.
Talk a little bit about IoT data and BI
Bradley: I have been talking for years about the IoT value formula being the intersection of Things, Data, Process, and People. What is a dark asset? What is something that is not connected in your process that can create value? How will you capture the data once that device is connected? When you apply machine learning/perform analytics on the data, what process will change? Finally, big data is worth nothing without big judgment. How will you get people to take action as a result of the analytics? Only after you have answered all these questions will you have a viable and valuable IoT solution.
Burton: The power of the Internet of Things stems from its ability to generate large volumes of granular data from powerful, inexpensive, IP-enabled sensors deployed across the systems customers need to manage. The problem is that companies tend to focus too much on data collection and too little on deriving actionable insights to benefit the business. To do this, look for solutions that identify patterns, trends, and exceptions; apply business logic; use the logic to kick off appropriate workflows; and optimize system performance over time. For example, think of a food company with multiple processing facilities. An effective IoT system would use sensors to monitor process temperatures across each facility; controllers to manage temperatures to ensure quality, safety, and compliance; and process intelligence tools to analyze performance across multiple factories, identify best practices, and apply them across all facilities.
Kern: IoT is a real opportunity. Again, BI and even AI working in concert with IoT sensors are already dramatically changing businesses from aviation to healthcare. As sensors become better and smaller, these technologies will be embedded in the technology ecosystem, enhancing efficiency, improving safety and increasing reliability. And we are just at the beginning.
Kogan: Connected devices and the Internet of Things are generating more data than we’ve ever seen before, but harnessing that data has been a challenge for most companies. We’ve seen customers using big data analysis combined with the IoT to improve employee productivity, accurately track and reroute package deliveries, and monitor large-scale manufacturing operations for efficiency and safety.
Orcutt: Data is everywhere and now that every device seems to be connected and generating information, systems can get overwhelmed. But this is a great thing in my mind. The explosion of information is incredible. I can recall years ago customers asking about using our advanced text analytics and analysis tools on telemetry use cases. Now with IoT, this is a no-brainer. People expect to understand so many different dimensions of information generated by processes, systems or just smart devices that are now connected and interacting with data services. Because of IoT growth, BI tools will be pushed further and have to become faster, more elastic/scalable, and smarter to help users make better decisions.
Describe the role you’d like to see for a Chief Data Officer
Bradley: I firmly believe a successful Chief Data Officer needs to be the digital change agent for their organization. In order to achieve this, there are three fundamental leadership principles they should embrace: 1) Build a listening organization. To be successful, companies must innovate outside the four walls of their organization. The CDO must provide a platform that allows the organization to listen, learn, and leverage resources and data that exists within and outside the organization. 2) Create a high integrity environment. Your data scientist team will tell you time and time again, Garbage In, Garbage Out. The value derived from machine learning and business analytics is directly proportional to the correctness of your data.
3) Build an inclusive organization. Instilling a diverse and inclusive culture that promotes full participation of all involved to maximize the sharing of information and driving towards fact-based decision making.
Orcutt: Data is a new currency or intellectual property. The role of the Chief Data Officer is absolutely critical in my mind. This person enables and exposes the required information that will change how businesses operate, engage customers, create new services, and ultimately succeed. This role is aligned closely with the strategy of the CEO and ensures that the innovation that companies require to compete in today’s real-time, hyper-engaged business is driven by decisions, insights and information contained within multiple data sources, documents, and repositories that is ultimately made available in a meaningful way to key stakeholders.
At what point does BI cross into cognitive computing?
Kern: The question of BI becoming cognitive is a tough one. As we look at today’s sophisticated AI, it is combining sophisticated software and powerful cloud-based computing. Cognitive takes it to another level where more contextual or even emotional awareness will become more sophisticated than today’s AI. I think it will happen based on the trajectory we are on today with BI and AI, and at some point we will see embedded cognitive device level applications.
Orcutt: I think both technologies are very complimentary. Cognitive computing uses AI and machine learning to deliver and understand context, relationships, and more complex data that can be made available to BI tools and systems to make the information meaningful to users. Cognitive tools are driven by learning and training of systems to automate previously manual or human tasks. While BI tools have similar approaches to understanding data, cognitive is focused on the deep learning and AI based technologies that deliver meaning, context and relationships in a smarter more scalable methodology. The benefits of cognitive computing exposes even more information, context, and relationships that could drive broader adoption of BI tools and systems.
This article originally appeared in the May 2017 issue of Workflow.