What is Embedded Analytics and How Does it Work?

As data takes centerstage for many organizations, business leaders are figuring out how to analyze that data as a means to business sustainability, agility, and competitiveness. Many organizations use traditional enterprise reporting solutions to make informed decisions, where the user views and interacts with dashboards and visualizations using a business intelligence (BI) tool.

Where these solutions fall short, however, is when users have to switch between their workflow interfaces and data tools to view the data and apply insights to the job they are performing. Data consumers need the ability to integrate data and analytics within their workflows so that they can take immediate action on the insights they glean.

Embedded analytics does just that. The streamlined nature of embedded analytics makes data-driven decisions more intuitive across the organization and brings potential for both internal and external users to get more value from company data.

What is Embedded Analytics and How Does it Work?

Embedded analytics places analytics at the point of need—inside of a workflow or application—and makes it possible for your users to take immediate action without needing to leave the application. Because it drives efficiencies within workflows, it increases user adoption of your analytics investments.

Think about an operational system like an ERP application and how business users traditionally use analytics to guide decisions. A business user may see low inventory on a dashboard in an analytics application, then they would need to open a new, separate application—the ERP—to put in the order for additional inventory. With embedded analytics, the business user can see the analytics and place their order within the same application. It’s not just the ability to see the information that brings value, but more so the functionality to immediately act on it in the same place.

What Are the Use Cases of Embedded Analytics?

Embedding analytics into your enterprise software applications can help speed up decision-making, streamline operations, improve customer relationships across the enterprise, and even be a new source of revenue.

Here are some common use cases for embedded analytics:

Increase Efficiency: Business users want to glean analytical insights in context of their workflows because it enables them to act fast and add more value. Embedding analytics brings the point of action into the workflow (no more toggling between applications) and enables the business user to solve a problem or seize an opportunity without leaving the application and interrupting their workflow.
Increase User Adoption of Analytics: When it’s easier for the user to take action with data because it’s integrated into their workflows, they are more likely to utilize company analytics and add value to the organization.
Reduce Error: Putting analytics side-by-side with mechanisms to take action dramatically reduces likelihood of user error that can occur when viewing and interpreting data in one application and applying the learnings in another application.
Create a Consistent Look and Feel Within Applications: Whether to satisfy branding requirements or provide familiarity to business users, having a user interface that looks and feels like other systems within your organization creates a seamless and approachable experience for everyone (and promotes user adoption).
Optimize Mobile Workforce: It can be extremely difficult and time-consuming to toggle among browser tabs and applications on a mobile device. Embedded analytics helps increase productivity by providing interactive functionality on the mobile device within a single application.
Address License Cost Constraints: One of the biggest impediments to company-wide use of analytics is the cost associated with granting access to it. Embedded analytics can be a way to avoid expensive license costs for users who do not need the full functionality offered in an analytics tool’s native interface.
Monetize Analytics: While data and analytics provide clear value to organizational decision-making, with embedded analytics, you can monetize your data by offering access to your customers. This could be done by embedding visualizations and dashboards into external-facing applications or portals or even providing a second tier of advanced analytics for customers at a cost.

No matter your business, if you want to do more with your data, embedded analytics can bring value to your organization in many ways.

What Are Some Real-World Embedded Analytics Examples?

Placing analytics at the point of need and making it possible for your users to act on insights your data can provide takes you one step closer to realizing the value of your data. Here are some real-world examples of how exactly that is done and the ways in which each company was able to make more informed decisions specific to their use case.

Marketing Visualization with Critical KPIs: The marketing team within Analytics8 was looking to embed Power BI visualizations on its website, which is hosted on a WordPress Platform. The team wanted to showcase examples of interactive dashboards created for retail customers—using data points from multiple sources—to show users how you can use these dashboards to quickly access critical KPIs and get a holistic picture of operations. With some simple coding, the marketing team was able to embed the dashboards on the Analytics8 website, making it possible for any external users to interact with the dashboards.
Focused, Customer-Branded Reporting Hub: A CPG company we worked with had built a custom SharePoint site for its warehouse managers to pull Excel reporting for time entry, PDF documents, and a number of other data points. To save time and reduce errors, we built an embedded analytics application utilizing Power BI and embedded the reporting hub directly into SharePoint so that managers could glean quick insights on overall warehouse performance in addition to the documents that already existed within SharePoint.
Retailer Management: Anheuser-Busch (ABI) wanted an application that managed all the retailers that sell ABI beverages. Within this app, ABI was looking to capture data around customer satisfaction—from start of delivery to next engagement—and to record real-time feedback so that they can better manage the relationship. We built an application via the Writeback Extension in Qlik Sense, which provided ABI the ability to quickly view—in real time—any feedback that is inputted and then take corrective action where needed.
Internal Mashups for Performance Reviews: A team of executives at Syngenta meet once a week to do an overall health check of the entire business. To get an accurate view of everything that was going on, they needed to pull information from across several different applications, and the solution they had in place was to take screenshots from different Qlik Sense applications the day of the meeting and piece it together in a PowerPoint presentation. Not only was this an inefficient use of time, but the team of executives could not interact with data that was being presented—they could not drill down or filter or look at different scenarios for what the data was telling them. The solution was to create an embedded analytics application within Qlik Sense that pulled data from across multiple applications in real-time. The team of executives have access to more valuable data, within one user interface, and can dedicate more time to understanding the value behind those insights.

How to Get Started with Embedded Analytics

Your embedded analytics solution requires more than the right technology; you must also consider the processes and people it will take for successful implementation and maintenance.

1.) Start by identifying your business goals. What problem do you want to solve, or what opportunity do you want to pursue? This will define the direction and scope of your project.

2.) Understand how embedded analytics scales with and affects your business intelligence (BI) tools. You need to assess your infrastructure, existing technologies, and resources to ensure you have servers and databases that can support an embedded analytics application. Make estimates up front to determine increased load on your BI tool and downstream databases. Assess the expected number of users, the frequency in which data will need to be updated, and the anticipated SKUs/capacity needed to host your embedded solution.

3.) Evaluate the technology you’re using to deploy embedded analytics. Even with so many technologies in the market that make embedding analytics easy (Looker, Power BI, AWS QuickSight, Qlik, Sisense, and countless others), you still need take the time to ensure you’re picking the best technology for your business and use case. You need to think about:

Skills Needed: Do you have the expertise needed to deploy embedded analytics? Typically, in order to ensure a successful embedded analytics deployment, you need people with web development skills, including skills to handle authentication and single sign-on, as well as traditional analytics skills such as data modeling and fundamental UI design.
Type of integration: Traditionally, people think about integrating analytics inside of an operational system. For example, an organization can place embedded analytics using Looker into its Salesforce dashboard so that its sales team can act on information immediately. That’s one way to do it. Another way to approach embedded analytics is to put operational capabilities inside of an analytics dashboard by creating a call to action within an analytics application. For example, an organization can place a button in a dashboard (that “speaks” to an ERP via APIs behind the scenes) that allows users to order new inventory.
Type of application: You can embed analytics into an existing application, or you can build a custom application from the ground up. With either of these use cases, you need to consider the look and feel of the software and the skillsets necessary to support development. You also need to consider how easy it is to deploy the product, how easy it is to scale, and how easy it is to understand and manage the licensing agreement for it.

4.) Plan for security vulnerabilities: Embedding data may expose security concerns, so it is paramount to address who should have access to what, how access is granted, and what users can do with it once they have it. When you integrate two or more systems, you’re creating a vector that may allow for unauthorized access. To ensure security within your analytics, consider these two things:

Authentication: If your goal is to reduce friction, then you don’t want users to have to provide logins for multiple behind-the-scenes applications; you want single sign-on (SSO). But SSO for embedded analytics can be challenging if you don’t address authentication at the beginning. You need to have an authentication strategy in place before beginning web development to decide which application is going to drive login requirements.
Access: Often data should be filtered differently depending on the end user. A strategy for data access levels should be established before development begins. This strategy, often coupled a data governance program, will allow you to define who should have access to what and when.

5.) Ensure your data and analytics workloads are migrated to the cloud: The cloud allows you to do more with embedded analytics through expanded API integrations, which makes complex uses cases such as write-backs and enhanced collaboration much easier to perform. A cloud infrastructure also allows you to scale as needed to support your embedded analytics applications.

When implemented with careful consideration of your business needs and analytical environment, embedded analytics will empower your business users to take action and enable your customers to get more value out of your data.