In fact, BI is alive and well; it’s becoming easier to use, it’s expanding out to more employees, it’s moving to the cloud, it’s becoming embedded in broader ERP and CRM software suites, and it now encompasses AI and machine learning.
According to 2020 market share numbers from IDC, the total worldwide business intelligence and analytics market hit $19.2 billion, growing a healthy 5.2% despite pandemic-related economic upheavals. Looking ahead, BI growth is expected to accelerate as companies focus on digital transformation and smarter ways to use data to drive the business forward.
The market leaders are a who’s who of the most powerful companies in the industry — Microsoft, SAP, Salesforce, IBM, SAS, and Oracle. On the other hand, those leaders only account for around 60% of the market, so there’s plenty of room for innovators such as ThoughtSpot and Alteryz to build up a following.
Still, BI has been around forever and one could argue it has been a bit of an underachiever. The knocks against BI have been that it’s too hard for the average worker to use, that it spits out fancy reports and offers up colorful dashboards that don’t help employees solve real-world business problems, and that it requires too much upfront work — creating data catalogs, building data warehouses, and so on.
Carsten Bange, founder and CEO of BARC Research Center, says that prior to the pandemic, BI was viewed by some as a legacy technology that in many cases had not proved to be worth the investment. That has “dramatically changed,” says Bange. New survey results show that companies are shifting their attention to BI once again, as they recognize the need to gain a deeper understanding of their supply chains, of rapidly shifting consumer behavior, and of their own business processes.
“Data-science based analytics is a high priority,” among companies surveyed in BARC’s Data, BI and Analytics Trend Monitor 2021 report, he adds.
Here are some of the key trends in BI for 2021 and beyond.
1. AI and machine learning offer exciting possibilities
The most significant trend in BI is the integration of AI and machine learning. “A new era of augmented analytics” has begun, declares IDC analyst Dan Vesset. “The AI-enabled analytics functionality that is required to bring this new generation of BI software to the masses is still nascent, but historical trends suggest that it will take less than 10 years for this generation of BI software to reach mainstream adoption.”
Boris Evelson, a Forrester Research analyst, adds that augmented BI (classical BI augmented with AI) has the potential to “turn the average business user into a citizen data scientist.” The goal is to enable non–data scientists to do forecasting, predictive analysis, anomaly detection, and other BI-related functions “with a single click,” according to Evelson.
In addition, machine learning systems can run in the background and address the problem of “not knowing what you don’t know.” Machine learning systems can identify interesting patterns in the data and alert the end user in a way that could never be accomplished otherwise, says Evelson.
Adds Bange, “Augmented analytics describes features that supplement human capabilities with machine learning to couple creative problem solving with unrivaled pattern recognition to get the best out of both worlds. The major goal is to make analytics and BI easier to use, to lower the entry barrier for casual users and at the same time increase the efficiency and effectiveness of power users.”
2. Cloud adoption surges in post-COVID world
Cloud adoption of BI software has been a trend for quite some time, but has certainly accelerated due to the pandemic, which forced employees to work from home and forced IT to provide remote access to key business applications.
Bange says that 50% of new BI deployments are in the cloud, which represents a steady year-over-year increase. The advantages of cloud-based BI include accessibility for remote users, scalability, elasticity, and speed of deployment. In addition, as companies become more comfortable moving large data sets to the cloud for backup purposes and for running applications, they are more likely to shift data warehouses and data analytics to the cloud. “Analytics leaders prefer to bring the analytics to the data, and not the other way around,” says Bange.
3. Natural language processing takes a step forward
Unless you’re a data scientist, formulating the right query can be difficult. The answer is building natural language processing into BI systems, so that the average employee can simply ask a question and get an answer. Natural language processing not only enables existing BI-trained employees to make better use of BI tools, but it enables companies to extend BI deeper and wider across the organization.
While natural language processing is certainly an interesting trend, it’s also fair to say that it’s not totally there yet. “Translating natural language into a precise query can be very challenging,” says Evelson. “You don’t always get the right answer on the first try. Maybe you get hundreds of answers,” similar to what happens when you do a Google search. Natural language systems still require quite a bit of tuning, he adds.
4. BI becomes embedded into CRM and ERP platforms
Whether through acquisition or internal development, the CRM and ERP vendors are embedding BI into their platforms. For example, Salesforce bought BI leader Tableau in 2019 and quickly integrated the software into its cloud-based CRM platform.
The advantage, according to Vesset, is that BI evolves from being a separate, disconnected process to becoming an integral part of the business process workflow. Embedded BI can help companies automate the steps involved in a business process, which delivers increased speed and better performance.
5. New ways of presenting information through storytelling emerge
In traditional BI, the system spits out reports and dashboards chock full of colorful charts, but that presentation, while slick, might not be the best or most useful way to present information to nontechnical users. Bange says a countertrend to the “highly sophisticated visuals” is a shift toward storytelling rather than data dumping.
Using the principles of a discipline called “information design,” BI vendors are simplifying their presentations in a way that walks the user through a particular problem or situation and not only presents raw data, but also provides recommendations for what to do. This type of narrative is more likely to include a textual narrative to accompany all the glitzy imagery.
6. BI becomes operational
Classical BI delivers reports on a fixed schedule, weekly or monthly for example. But that’s no longer enough in today’s competitive business environment in which decisions need to be made in real-time. With operational BI, also known as operational intelligence (OI), data from various sources, including consumer behavior and supply chain disruptions, is collected and analyzed.
The BI system is then able to provide recommendations for quick decisions, such as assigning more resources to a specific function or responding to a fast-changing business condition, says Bange. With operational BI, dashboards can automatically refresh at fixed intervals, such as every hour, and the system can trigger alerts to notify operational teams that there’s a problem that needs to be addressed, or an emerging opportunity that can be exploited.
7. Successful BI continues to require upfront work
BI tools themselves are well-established, but many companies struggle to implement BI because they have not done the necessary prep work. “The technology is mature,” says Evelson. The hurdles exist on the people and process side of the equation. Companies need to build a data-driven culture. They need to train employees.
According to the latest survey research from BARC, when respondents were asked to rank their 2021 priorities, data quality management and data discovery topped the list. Advanced analytics and machine learning ranked 11th, which doesn’t mean that companies aren’t interested in AI. It means that “companies are struggling to adapt machine learning mechanisms when the foundation — good quality and accessible data — has not quite been achieved.” Bange says that companies “seem to be going back to the roots and concentrating on the basics of using and managing their data before they shift priorities onto advanced methods.”
Evelson’s recommendation for CIOs is to “get on an enterprise-grade platform right away,” whether that means refreshing an older version of a current BI platform, or going with a new vendor. He points out that only between 20-30% of the data that could be used for analytics is currently being pulled into the average enterprise data warehouse. BI is “an investment in everything that a CIO needs to have in order to be successful,” he adds.
Author: Neal Weinberg