Business intelligence (BI) software has gained considerable traction since its introduction as "decision support systems" in the 1960s. Today, there are over 100 BI software companies selling some type of business intelligence tool. We put together this buyer's guide to help buyers understand the market. In this guide, we'll review:
BI software helps organizations organize and analyze data to make better decisions. This could include internal data from company departments as well as from external sources, such as marketing data services, social media channels or even macroeconomic information.
The BI market is growing rapidly because of the proliferation of data to analyze. Over the past few decades, companies that have deployed Enterprise Resource Planning (ERP), Customer Relationship Management (CRM) and other applications are now sitting on a mountain of data that can be analyzed. In addition, the growth of the Web has increased the demand for tools that can analyze large data sets.
One of the biggest trends in the BI market is the shift in software architecture and design to more user-friendly applications. These applications are now being used by business users—not just IT staff—to analyze particular sets of departmental data, including marketing, procurement, retail and Web data.
BI software can be divided into three broad application categories: data management tools, data discovery applications and reporting tools (including dashboards and visualization software). In the next section, we'll explain how these applications can help your organization's decision-making process become more data-driven.
What BI tools you need depends on how your data is currently managed and how you would like to analyze it. For example, if it is currently scattered across disparate transactional databases, you might need to build a data warehouse to centralize it and invest in data management tools that offer Extract, Transform and Load (ETL) functionality to move and re-structure it.
Once data is given a common structure and format, you can invest in data discovery solutions such as Online Analytical Processing (OLAP), data mining and semantic or text mining applications, with the capability to create custom, ad hoc reports. And because information is stored within the warehouse, users can quickly pull reports without impacting the performance of the organization’s software applications, such as CRM, ERP and supply chain management solutions.
We’ve illustrated this concept in the image below:
But this isn’t the only way to implement BI within your organization. If you’re only analyzing data from a single source, ETL and data warehouses are unnecessary. Alternatively, you might require multiple warehouses, and thus, require different tools to connect data between both these servers and other BI applications that need access to this data.
Better decision-making starts with better data. Data management tools help clean up "dirty data," organize information by providing format and structure and prepare databases for analyses.
|Data quality management||Helps organizations maintain clean, standardized and error-free data. Standardization is especially important for BI implementations that integrate data from diverse sources. Data quality management ensures that later analyses are correct and can lead to improvements within the business.|
|Extract, transform and load (ETL)||Collects data from outside sources, transforms it and then loads it into the target system (a database or warehouse). Because primary data is often organized using different schemas or formats, analysts can use ETL tools to normalize it for useful analysis.|
The ability to sift through data and come to meaningful conclusions is one of the most powerful benefits of adopting business intelligence tools. Data discovery applications help users make sense of their data, whether it be through quick, multivariate analysis during OLAP or via advanced algorithms and statistical computations during data mining.
|Data mining||Sorts through large amounts of data to identify new or unknown patterns. It is often the first step that other processes rely on, such as predictive analytics. Databases are often too large or convoluted to find patterns with the naked eye or through simple queries. Data mining helps point users in the right direction for further analysis by providing an automated method of discovering previously neglected trends.|
|Online analytical processing (OLAP)||Enables users to quickly analyze multidimensional data from different perspectives. It is typically made up of three analytical operations: data consolidation, data sorting and classification ("drill-down") and analysis of data from a particular perspective ("slice-and-dice"). For example, a user could analyze sales numbers for various products by store and by month. OLAP allows users to produce this analysis.|
|Predictive analytics||Analyzes current and historical data to make predictions about future risks and opportunities. An example of this is credit scoring, which relies on an individual's current financial standing to make predictions about their future credit behavior.|
|Semantic and text analytics||Extracts and interprets large volumes of text to identify patterns, relationships and sentiment. For example, the popularity of social media has made text analytics valuable to companies with a large social footprint. Understanding semantic trends is a powerful tool for organizations evaluating purchase intent or customer satisfaction among users of these channels.|
In the words of John W. Tuckey, “the greatest value of a picture is when it forces us to notice what we never expected to see.” Reporting applications are an important way to present data and easily convey the results of analysis.
BI users are increasingly business users—not IT staff—who need quick, easy-to-understand displays of information. In response, software vendors have been working to mask the complexity of these applications and increasingly focus on the user experience.
|Visualizations||Helps users create advanced graphical representations of data via simple user interfaces. The ability to visualize information in a graphical format (as opposed to words or numbers) can help users understand data in a more insightful way. In addition, new interactive tools can provide teams the ability to both analyze and manipulate reports in real-time.|
|Dashboards||Dashboards typically highlight key performance indicators (KPIs), which help managers focus on the metrics that are most important to them. Dashboards are often browser-based, making them easily accessible by anyone with permissions.|
|Report writers||Allows users to design and generate custom reports. Many CRM and ERP systems include built-in report writing tools, but users can also purchase standalone applications, such as Crystal Reports, to create ad hoc reports based on complex queries. This is especially helpful for organizations that continually modify analyses and need to generate new reports quickly.|
|Scorecarding||Scorecards attach a numerical weight to performance and map progress toward goals. Think of it as dashboards taken one step further. In organizations with a strategic performance-management methodology (e.g., balanced scorecard, Six Sigma etc.), scorecards are an effective way to keep tabs on key metrics. For example, a scorecard might establish a grade of “A+" to 40 percent year-over-year growth if the goal was set at 14 percent.|
Before evaluating software, you must determine what type of buyer you are.
Business users and departmental buyers. These buyers favor small data-discovery vendors and BI tools over the big, traditional BI systems. Ease-of-use and fast deployment are more important than in-depth functionality and integration. They are usually business users rather than IT staff.
IT buyers. Traditional buyers are more focused on functionality and integration within their information infrastructure stacks or other ERP applications. Integration across different entities and departments is usually more important than ease of use.
As you begin your software comparison and evaluation, there are a couple trends to consider:
In-memory processing. OLAP systems of the past would pre-calculate every possible combination of data. These calculations would be stored in the “cube,” and users could retrieve them when they needed a certain analysis. Creating these cubes was very time-consuming—sometimes taking as long as a year—and required expertise. Today, computer processors and memory are faster, cheaper and more powerful overall. This same process can happen in-memory, rather than using a disk-based approach with cubes. Analytic software built on an in-memory architecture can retrieve data and perform calculations in real-time or on-the-fly.
Big Data. The Internet is rapidly creating vast amounts of data. This phenomenon is being dubbed “Big Data” among IT and business leaders. Business analytics software companies are beefing up their data warehousing and analytics capabilities to keep up with demand.
However, according to Gartner, through 2015, 85% of Fortune 500 organizations will be unable to exploit big data for competitive advantage. The right BI tools can help harness the power of so much data.
Companies dealing with large amounts of data may also want to consider investing in dedicated IT security suites to support their computer security needs.
Business users to outnumber IT staff. This is a major trend playing out in the market. More business users—rather than traditional IT staff—are evaluating and purchasing software. So usability is becoming more important than functionality during software evaluations. As a result, small data discovery vendors that develop really good interactive visualization tools are gaining market share. Meanwhile, traditional BI vendors are parroting new market entrants by promoting ease of use.
Software-as-a-Service (SaaS). A growing number of organizations are considering SaaS or “cloud” BI software instead of traditional, on-premise software that you install on-location. Cost is a major driver of this trend. The poorly performing economy is motivating companies to look at lower-cost BI software from SaaS and open source vendors. Of course, perceived ease of use, faster implementations and reduced IT needs are also driving this trend. On-premise BI vendors are responding by committing development resources to cloud technology.
Mobile BI applications. Proliferation of the iPhone, iPad and other mobile devices is pushing vendors (e.g., Microsoft and Oracle) to develop on-the-go business intelligence applications. Analysts think mobile BI could expand the population of BI users to a larger, mainstream audience.
Microsoft acquires Datazen. In April 2015, Microsoft announced the purchase of mobile BI platform Datazen. The move is part of Microsoft’s ongoing initiative to improve the functionality of Power BI on mobile devices.
Birst raises $65 million. Cloud-based BI vendor Birst raised $65 million in a venture round as it ramps up for an IPO.
Hadoop adoption still limited. In 2015, Gartner reported survey results suggesting that adoption of Hadoop, an open-source programming framework used in the storage and processing of very large data sets (“big data”), is still limited, with only 26 percent of respondents claiming to be currently deploying or experimenting with the platform.
We're able to offer this service to buyers for free, because software vendors pay us on a "pay-per-lead" basis. Buyers get great advice. Sellers get great referrals.