Create, deploy, and maintain analytic applications that engage users and drive revenue. See a Logi demo

BI Trends

Embedded Analytics Maturity Model: It’s not what you embed, it’s how you embed it

By Alvin Wong | August 14, 2014

In the realm of human psychology and communication, there is a saying: “It’s not what you say, but how you say it.” Meaning the tone of a speaker’s voice and the non-verbal aspects of a presentation influence the message being conveyed much more than the spoken words themselves. The key to delivering a message lies in components such as tone of voice, maintaining eye contact, body posture, and facial expression.

Today, organizations want to be more data-driven, which often requires embedding analytics inside the applications their employees use every day. As it turns out, how you embed these capabilities into an application affect the user experience and the business outcomes just as much as deciding upon what capabilities are embedded.

In this post, we will look at how analytics are integrated into software applications, examine the key integration points, and see how infusing analytics into the core workflows of an application improves the user experience and creates a more competitive product.

Embedding = Integration

Embedded analytics is the integration of analytic capabilities inside business applications. It is often presented to users in the form of dashboards, reports, and self-service analysis functionality inside applications such as a CRM, a financial application, or electronic health records systems. When integrating functionality, there are four areas of the application to integrate with.

  • Data – analytics is driven from data, so connect to the data your product or application generates; think about the data to be utilized in the future, not just today; users often require real-time access to data
  • Security – if your application environment has a security model and users already know how to log in, leverage that as much as possible instead of creating yet another set of credential users need to memorize
  • User interface – embed analytic functionality into your application so it looks and acts like a part of your application
  • Workflow – operationalize analytics by integrating into the core workflows and functionality of your application

Infused Analytics

What makes a great user experience for consumer applications today is the way they incorporate data and analytics as a natural part of their application, providing the necessary information that leads to your desired transaction. For example, Amazon provides product reviews and suggestions alongside the Buy Now button. Real estate apps like Zillow provides map-based searches that leads to your next mortgage application and contact with a real estate agent.

By infusing analytics directly into core application workflows like consumer applications do, business applications can also create a powerful, innovative user experience, by merging analytic insight with their transactional capabilities. For example, a CRM can provide a sales manager the ability to analyze sales performance by territory, identify a poor-performing territory, and change territory assignments on the same screen. Another example is an insurance claims application where a claims examiner can interactively compare a claim against similar claims within the approval workflow.

According to our survey of software and SaaS providers, those who infuse analytics into their core workflows most strongly believe that analytics helps them to improve the user experience, differentiate their product, and increase revenue.

Embedded Analytics Maturity Model

Infused analytics is not the only way to incorporate analytics into a product or application. Many developers are not ready for this level of integration, or it may not be appropriate for the application. The Embedded Analytics Maturity Model outlines the four ways of integrating analytics with an application, each with progressively deeper integration.

At the beginning of the maturity model, there is a product or service that generates data and the standalone analytics application enables a user to understand the data. What is most important for that separate application is to access the data. In commercial applications, this model is often employed when the standalone analytics application is the only user interface offered. For example, consider a service such as Google Analytics where a central repository of data is fed by website interactions, and the site administrator logs into a web application to view the reports on web visitors.

In the first stage of embedding, the security model is integrated between the business process application and the analytics application. Users log into the business process application they use every day, and click on a link to go the analytics application, where single sign-on has been enabled. In this model, the access to analytics in embedded. The rights and roles configured in the process application are passed to the analytics application and controls the data the user has access to. This model is often employed with hybrid architectures, where the analytics application is hosted in the cloud and reports against data coming from one or more on-premise applications.

Inline analytics builds upon the first two models and adds integration at the UI or presentation tier. This places analytic content and capabilities inside the business application. Analytics is available to the user in the reports module of the application, or as a dashboard on the homepage. From an interaction standpoint, a user can usually click on a link from a visualization to an existing page in the application.

What are the latest trends in Embedded Analytics? Download the 2016 State of Embedded Analytics Report

With infused analytics, the application is developed with analytics as a core capability and woven into the user workflows. Analytics is integrated at the application or business logic tier. Analytics is embedded within workflows to guide user action – knowing and doing is available in the same context instead of separate contexts.

The Future of Embedding

While we have laid out four different ways applications embed analytics, software providers should consider which is best for them, their customers, and for their market. In our survey, we saw that inline analytics being the most popular and infused analytics second. However, software providers who have embedded analytics in some form today anticipate that they will integrate analytics deeper into their applications in the future.

To learn more see the full Infographic or download the report now.


About the Author

Alvin Wong has an extensive background in solution architecture and implementation of SaaS and business intelligence applications. Alvin earned his MS in Engineering Management from Stanford University and BS in Electrical Engineering from Cornell University.

Subscribe to the latest articles, videos, and webinars from Logi.