Business Intelligence Tools:
A Comparison Guide

To say that today’s market is flooded with new business intelligence tools might be a bit of an understatement. Gartner’s 2016 Magic Quadrant for Business Intelligence and Analytics reviews 24 of the leading vendors, and there are still dozens more that did not make the cut. This is only natural, considering that this is a new and lucrative market that’s estimated to be worth around $17 billion.

While it would be virtually impossible to review each and every tool (although valiant attempts have been made), this guide is meant to serve as an overview that will help you understand the marketplace and the types of existing products, as well as the specific use cases they are best suited for. Hopefully, this will allow you to better navigate within the crowded marketplace and find it easier to classify the various tools you encounter.

What are Business Intelligence Tools?

The first thing that is important to understand is that business Intelligence, as a field, is relatively new, and the terminology surrounding it is quite muddled (for evidence of this, try getting a straightforward explanation of the difference between ‘business intelligence’ and ‘business analytics’). TechTarget offers a concise definition of BI:

Business intelligence (BI) is a technology-driven process for analyzing data and presenting actionable information to help corporate executives, business managers and other end users make more informed business decisions.

Accordingly,data analytics tools are basically computer programs that facilitate the process of analyzing data and presenting the results of data analysis to stakeholders within an organization, in order to promote data-driven decision making, corporate transparency, or other goals.

The Implications of Data Complexity

Before we approach the matter at hand, another introductory note must be made regarding the matter of data complexity, which is often the dividing line between the various available BI tools. While most of these softwares will produce satisfactory results when pitted against simple datasets, there are certain unique challenges regarding complex data.

As any data professional might tell you, the process of analyzing complex data – which in this context would mean large and diverse datasets from multiple data sources – often contains several stages, with a lot of the work actually going into the initial stages of extracting, cleansing and modeling the data before any actual analysis or visualization can be performed. Accordingly, there are tools that address parts of this process, often making them best suited for simple data but problematic for working with complex data; while full-stack and single-stack™ platforms aim to provide an integrated solution for more difficult data scenarios.

The bottom line for this section is that it is imperative to understand the complexity of your data before you begin to investigate BI solutions. A tool that does wonders for a company with simple data could lack core functionality needed by an organization that works with big or disparate data.

The Seven Types of Business Intelligence Tools

The vast majority of business intelligence tools can be classified in one of these seven categories:

(1) Spreadsheet Apps

Required technical skills: Low
Price: Very low
Flexibility: Medium
Supported data types: Simple, small

Excel or similar apps such as Google Sheets are by far the most common and basic BI tool, and you’ll be hard pressed to find an organization that doesn’t use them. Indeed, they offer unparalleled ease of use for tabular calculations and quickly applying simple mathematical formulas. However, organizations that have more serious ambitions about their data often quickly outgrow these types of applications, as they tend to be bound by data size limitations, either in hard maximums or performance issues, and offer limited capabilities in terms of data preparation, reporting and visualization. But at a price tag of virtually nothing (in enterprise software standards), these tools will still be found in almost every type of company, regardless of size.

(2) Traditional Enterprise BI

Required technical skills: High
Price: High
Flexibility: Low
Supported data types: All

Legacy business analytics tools have been around for decades, and usually offer a highly robust, all-inclusive solution for enterprise-wide warehousing and analytics. The analytical functionality is often offered as part of a wider suite of products, with various modules needed in order to perform tasks such as native data imports, ETL and dashboarding. While these tools are still pervasive in many larger organizations, they are notoriously inflexible, requiring large IT teams to operate and maintain on a day-to-day basis. While these types of platforms will be able to handle any data you throw at them, the reliance on IT resources, as well as limitations stemming from outdated technology, can often lead to major slow-downs in the time it takes to generate new reports, and provides very little aid in terms of self-service analytics for non-technical users.

(3) Modern Data Discovery

Required technical skills: Medium
Price: Medium
Flexibility: High
Supported data types: Simple

Since the mid-2000s the traditional legacy BI model is being replaced by more modern BI tools specializing in ‘data discovery’. These agile business analytics tools are easier to use and produce very attractive visual results, and are a perfect fit for organizations that work with simple data (i.e. moderately sized and well-structured), as they allow any user, regardless of technical skills, to both view and create very intuitive dashboard displays with a wide range of visualizations. However it should be noted that while these tools are often very easy and simple to use on the front-end, the back-end is another story – they will tend to offer very limited functionality when it comes to data preparation, modeling and ETL; or alternately they will abandon the simplicity of their front-end interface in favor of heavy scripting and coding. In addition, as these tools tend to rely on in-memory technology, they are fairly resource intensive and not well-suited for data that exceeds the few hundreds of millions of rows. Both of these issues are negligible when working with simple data; however companies that have complex data will often find that data discovery tools are only half a solution.

(4) Cloud BI

Required technical skills: Medium
Price: Varies
Flexibility: Medium
Supported data types: Varies between products; usually not a good fit for Big Data

Cloud-based business intelligence tools offer to host your data on the vendor’s infrastructure, removing the need to worry about server maintenance, hardware upgrades or version updates. This is an attractive solution for organizations that already work with mostly cloud-based data sources. There are various major cloud BI products available, each with different capabilities in terms of data complexity; however it is important that there is a price to be paid for leaving your data infrastructure in a vendor’s hands. You will often discover that adding new sources or data models is far from simple, and will require you to employ the BI provider’s professional services team, who is familiar with the Cloud infrastructure in which your data is being stored. Some industries, such as finance or healthcare, could also face regulatory issues around moving sensitive data to the Cloud. Finally, Cloud storage costs can easily skyrocket when working with larger datasets.

Read more about business intelligence in the cloud.

(5) Open-Source Business Intelligence Software

Required technical skills: High
Price: Varies
Flexibility: Medium
Supported data types: Varies

Open-source is not really a product classification – obviously, there can be many types of open-source products, which would fit into one of the above categories of business intelligence apps. However, since many smaller, more technical-oriented organizations tend to look for open-source solutions, we have included it on this list. These types of tools can usually provide basic functionality free of cost (for a company with existing programming knowledge). However it is important to note that for more advanced use cases, these solutions will often not be cost-effective – as the amount of internal or external development effort that needs to be invested in order to accommodate these use cases will often exceed the price of purchasing a non-open source solution.

(6) Data Preparation Tools

Required technical skills: Medium
Price: High
Flexibility: Medium
Supported data types: All

Another type of tool, that is very different than the ones mentioned above but is gaining popularity in recent years, is the proprietary data preparation tool. These offer little to no functionality in terms of front-end, and instead are focused on data cleansing, integration and modeling, often offering native connectivity to data discovery platforms. While the combination of these two would usually provide reasonable results for organizations working with complex data, data preparation tools tend to be on the costly side, and the need to also purchase front-end applications – as well as maintain and manage your ‘assembly line’ of data products – could easily lead to extremely expensive deployment.

(7) Sisense: Single-Stack Business Intelligence Software Built for Complex Data

Required technical skills: Low
Price: Low
Flexibility: High
Supported data types: Simple, complex, big

Sisense’s business intelligence tool is focused on simplifying business intelligence for complex data. This means giving users the agility of modern data discovery platforms, while avoiding the limitations these platforms usually faced in terms of data complexity. This is accomplished due to Sisense’s unique In-Chip™ data engine that processes more data while employing less computational resources, thus cutting down on many of the steps that would have typically been required during the data preparation process when working with complex data (as other systems would need to aggregate or flatten parts of the data in order to cut it down to manageable chunks). This further fits into the Single-Stack™ architecture, as Sisense is an integrated product that gives users everything they need to easily prepare, analyze and visualize data – even, and especially when working with large and diverse datasets. Sisense can be deployed on-premises or in the cloud.

Read more: Sisense compared to the alternatives