There are more players in the Business Intelligence (BI) space than ever before. For our clients, that means having so many options that it can be overwhelming. At Waterloo Data we are vendor agnostic and don’t have reselling relationships in place with any BI tools or any other vendors in the Data Management space. For this reason, we are frequently called upon to help our clients with vendor evaluations and vendor selection efforts.
Of all the BI products out there, the top three we most frequently encounter are Power BI, Tableau, and Looker. Each one of them has their pluses and minuses. In this post, we will compare and contrast these products based on real-world implementation experience.
Over the past few years, Microsoft has made major investments in Power BI resulting in displacing Tableau as the leading BI vendor in Gartner’s Magic Quadrant back in 2019. One of the more compelling reasons organizations adopt Power BI is that there is a “free” version of Power BI included with most Office 365 and Microsoft 365 plans. The “free” version is perfectly fine for individual analysts who want to work with their data, but you will need the “paid” version to share reports, dashboards, or analytical apps.
Pros of Power BI
The following are the primary strengths of Power BI:
- Price: Power BI is relatively low-cost when you look at all it offers. In fact, the entry and dominance of Power BI has essentially caused other BI vendors to become much more competitive in their pricing/licensing.
- Integration with the Microsoft ecosystem: The corporate world runs on Microsoft. Power BI integrates with Office 365 and Microsoft Teams. The Power BI cloud offering also works well with other cloud native offerings from Microsoft including Synapse and Azure Data Factory. It is also relatively straightforward to “embed” Power BI content (reports/dashboards) into internally developed .NET applications, thus bringing analytics directly to end users and not requiring them to go to a separate app or portal.
- Continued innovation: Microsoft continues to invest in additional visualizations as well as including AI and ML capabilities that leverage the capabilities of the Azure infrastructure.
Cons of Power BI
The following points are some of the primary challenges with Power BI:
- Gaps between on-premises version: In an attempt to drive more customers to their Azure cloud offering, Microsoft offers fewer features in the on-premises version of Power BI versus what they offer in the Power BI cloud service.
- Azure only: Microsoft does not support running Power BI as a cloud service on other platforms such as Google Cloud or Amazon Web Services.
- Content management: The content promotion and publication process can be clunky when working at an enterprise scale. Allowing self-service usage among business users can lead to a proliferation of Workspaces with little to no supporting governance capabilities. This is an area where the Power BI team at Microsoft is actively investing.
For years Tableau was the undisputed king of premium BI tools. They by-passed IT and sold directly to the business and Tableau Desktop became ubiquitous. Almost overnight, there was a proliferation of ungoverned analytics all throughout an organization. Business users loved the ease of use but struggled to publish, share, and disseminate their work. The product then evolved from a desktop version to a server-based version. The recent acquisition of Tableau by Salesforce has resulted in the ability to easily embed Tableau visualizations into Salesforce. Salesforce has also acquired MuleSoft and Slack for better data integration and collaboration capabilities.
Pros of Tableau
Here are some of the strengths of Tableau:
- User Experience: One of the most compelling features of Tableau is the overall user experience. Users can easily create reporting and dashboard content and use the tool for data exploration. From a UI perspective, Tableau is the standout among Power BI and Looker.
- Available Resources: Given the widespread adoption of Tableau it is relatively easy to find resources at all levels to support and use the product.
- Salesforce: Most organizations are using some aspect of Salesforce, having Tableau as part of the Salesforce ecosystem provides a solid BI tool to use across the Salesforce offerings.
Cons of Tableau
Here are some of the weaknesses with Tableau:
- Not cloud native: Tableau does not have a cloud native architecture and thus cannot take advantage of the elasticity of a cloud offering to automatically scale to support larger workloads.
- Pricing: Tableau remains one of the pricier BI tools in the market.
- Salesforce acquisition: Salesforce will continue to invest in new capabilities and integration within the larger Salesforce ecosystem, but the reality is that Tableau is no longer an “independent” BI tool and product development will have a Salesforce flavor.
Looker was acquired in 2020 by Google and has started to integrate into Google Cloud Platform’s (GCP) portfolio and go to market efforts, thus increasing market recognition. Prior to the Google acquisition, we would see Looker heavily marketed to the Snowflake community where they made big inroads alongside FiveTran. The Looker/FiveTran/Snowflake stack was a very popular choice for companies looking to jumpstart their data warehousing and BI efforts. Looker has recently flexed it’s innovation muscles by incorporating Natural Language Query processing, introducing a developer-friendly extension framework, and releasing Looker Data Dictionary. Looker’s integration within GCP makes it a great choice for companies that have adopted GCP as their platform of choice.
Pros of Looker
Looker has the following strengths:
- Performance and Governance: Looker uses the underlying power of the database to perform calculations and data processing. By relying on the power of the database to perform set functions and relational algebra, it delivers a snappy user experience, especially when running on top of a powerful MPP platform such as Snowflake. It has a powerful semantic layer (LookML) that allows developers to reuse data and calculations in a consistent manner.
- Extensibility: Looker is geared towards developers and provides extensive APIs, SDKs, and tools that allow developers to embed Looker into analytical applications, customer portals, and other customer-facing applications.
- GCP integration: Integration with Google BigQuery and the ongoing growth of GCP will make Looker a very attractive option for teams who are building their infrastructure around GCP.
Cons of Looker
Looker has the following concerns or risks:
- Data modeling using LookML requires programming: Some of the other vendors in this space are very much geared towards business users and providing self-service BI capabilities. Looker still requires coding to develop data models and isn’t as business-friendly as some of the other tools in this space.
- Not quite ready for full enterprise adoption: Where we see Looker the most is at SaaS companies that have a lot of engineering talent and are using Looker for dashboards and reporting. Its limited scope of enterprise capabilities such as collaboration, business-friendly self service capabilities, and other full-feature options would put it at a disadvantage to other products when being considered by Fortune 2000 companies. Looker also gets kinda pricey as you consider rolling it out across the whole organization.
We live in a time when we have access to some fantastic BI tools and resources. For most businesses and data professionals, it’s not a case of finding the ‘best’ BI tool. Instead, it will be all about finding the best fit. This will depend on the situation, the environment, and the goals of the data team.
Power BI is ideal for those teams heavily invested in MS, and want the drag and drop UI to get their insights. Tableau is for those that want presentation-level visualizations, in-depth analytics, and a dizzying range of features. Looker has a focus on performance and embedded analytics.
There is simply no ‘best’ option. The architecture and environmental factors are going to play a key role in choosing the BI tool for any business. The goal should always be to build your data landscape and architecture in a way that is presentation agnostic and allows you to support a bring-your-own-tool approach. By focusing on fundamental data architecture that is purpose built for analytics, you can support multiple BI tools with the same underlying data architecture.