An Introduction to Embedded Analytics

Billions of digital transactions are now taking place every day. Just think about it: we send messages all day, browse tons of websites, make online purchases,… And all these actions create valuable bits of information.

As a result, any company nowadays sits on a ton of data. Nevertheless, they don’t necessarily know how to go from raw bits of information to a data-driven decision making process. Unsurprisingly, the companies that do, are now emerging as the most competitive.

For this reason, more companies seek to unlock these insights out of the platforms and SaaS products they use daily. In this article, we’ll uncover how embedded analytics can support these companies in their digital transformation.

What is embedded analytics?

Embedded analytics is the integration of dashboards and other analytical features as an inherent part of another business application. This could be inside a software platform, a web application, an internal employee portal or even public websites.

In other words, companies using embedded analytics are essentially offering a window for the final user – or “data consumer” – to gather insights into the data stored within a specific software application or product.

Instead of keeping these insights in separated data platforms, companies can now include them as a core building block in their own business workflows. Typically, these insights are displayed as intuitive charts & dashboards on a broad range of user interfaces and devices.

Especially for technology facilitators like SaaS, IoT providers or service organizations, this is an absolute game-changer. Embedded analytics becomes a strategic differentiator towards their end-users, who explicitly look to consume and take action on these new insights.

What are typical use cases of embedded analytics?

While we typically think of data analysis in a B2B context, embedded analytics extends way beyond the business context only. To name just a few examples:

  • B2B Software-as-a-Service (or SaaS) companies who want to add intuitive reporting dashboards as an additional offering for their end-customers;
  • Enterprise organizations, using data dashboards as an internal information stream to collaborate between teams or departments & monitor performance;
  • Direct-to-consumer organizations who are looking to innovate interactions with their customers, members, employees or followers,…

To give some more specific examples of these direct-to-consumer organizations:


How are traditional BI and embedded analytics different ?

Traditional BI has been around for decades, and is a cornerstone to present-day business intelligence. However, there are a couple of crucial differences and drawbacks to the traditional BI approach:

  • Traditional BI is often complex and managed by a data analyst or IT, it’s less intuitive for business users;
  • They are stand-alone. It completely separates data analysis from the existing workflows and business processes. It forces users to switch between applications when they want to act upon the insights uncovered through traditional BI;
  • The lack of integration with other tools & business processes makes data analysis less accessible for most members of an organization;

Typical BI dashboards offer a centralized perspective of information, with extremely aggregated views across the entire organization. These views are usually owned by data scientists, but consumed by business users with little control over the analysis process.

Unlike traditional BI, embedded analytics gives more control to the actual business user who is in need of the insights:

  • Reporting dashboards can be firmly embedded into the core workflows and business applications of the final end-user;
  • The analysis & decision-making go hand-in-hand, as they happen within one and the same UI;
  • It unlocks information on a tactical level, allowing any member of an organization to immediately take action, driven by data;

To briefly summarize the biggest differences:

The growing market of embedded analytics

How come this relatively new phenomenon is sky-rocketing, even though traditional BI dominated the market in the last decades?

Well, the end goal of any embedded analytics project is to enable data-driven business decisions with the shortest time-to-insight. In order to achieve this for the end-consumer, two factors are highly important:

  • the degree of embedding: there’s a difference between embedding a single reporting dashboard inside a webpage, or having multiple charts interlinked with each other and with other workflows inside the core platform;
  • the degree of platform interactivity: the more interactivity options the end-user has, the more effective the analytics component will be. For example: access to advanced filters, drilldown on specific data, or triggering an event straight from the dashboard interface,…

In the past, many companies chose to develop charts and data visualizations in-house. However, a truly embedded and interactive solution requires considerably more budget, skilled resources, and development time before going to market. Therefore, the build-it-yourself model has proven uninteresting for fast-growing companies in the software, IoT and SaaS industries.

For these fast-growing technology sectors, off-the-shelf embedded analytics solutions are a more scalable investment than in-house solutions or traditional BI. They enable a faster go-to-market, without losing sight of the core product.

The rising demand for easy analysis

However, the success of embedded analytics goes beyond easy deployment. The ease of analysis is at least as important. Data has become a strategic asset for:

  • The data facilitators – typically technology vendors, SaaS platforms, IoT or service providers that are generating the business data;
  • The data consumers – the final users of these technology platforms, who want access to the data inside their business applications;

And for both the facilitators and consumers, embedded analytics solves some of the key challenges they face:

  • Data facilitators can now easily monetize data they are already collecting inside their platform. They can expand their offering with analytics add-ons;
  • Data consumers finally get easy access to data insights from within their core business applications. They can access the insights while they execute their day-to-day operations;

In short, the rising demand for self-service analytics and real-time visualization tools within business applications create new opportunities. Most market research studies predict growth rates for embedded analytics market between 10-15% CAGR for the coming years.

That leads us to the final important question:

What should a truly embedded analytics solution look like?

In essence, it comes down to the following:

“Embedded analytics tools should enable data insight and action in the relevant business area, through deeply embedding analytical features into the business model and core processes.”

Now, let’s break this down step by step. What are some of the must-have components for an off-the-shelf embedded analytics solution? We call them the 5 building blocks of embedded analytics.

Seamless integration

The first critical thing to consider for an embedded analytics component is the technical fit. In other words, how does the solution integrate with a product’s existing tech stack and architecture?

A couple of important things to assess are:

  • Flexibility: powerful embedded analytics building blocks are API-first. They allow for custom data connections, customized analytics functionalities, or other customizations to cope with changing or unpredicted business needs;
  • Low-code: search for solutions that allow deep integrations with an existing application’s workflows without writing endless amounts of code. Think of generating dashboard exports, executing queries, managing alerts, etc.;
  • Security: the option to re-use existing authentication and security mechanisms on the embedded dashboards is a must-have for many;
  • Scalability & Performance: the ability to easily automate personalization on a mass scale. Especially when serving thousands of users while balancing performance;
  • Technology-agnostic: an embedded analytics component should be embeddable in any technology. So, it shouldn’t matter which development frameworks someone uses (Angular, Vue, React,…), which data sources or which hosting services;
  • Responsiveness: reports and dashboards should be adaptive to the environment where they are displayed. Whether that’s on mobile, desktop, tablet or a large monitor;

Communication between dashboard & platform

To obtain a truly embedded analytics module, it’s essential that the analytics dashboard can communicate with the core functionalities of the platform in which they are being embedded.

Here are a few examples of what communication should look like:

  • Action triggering: the possibility for an end-user to trigger an action right from within a chart, while analyzing the data;
  • Automated alerting: let users set up automated notifications when the data meets a certain threshold, so they can immediately take action;
  • Collaboration: sharing data and insights across a user’s business applications, or sharing insights with team members;

A single user experience

Switching between business applications is the absolute worst for a user. For a state-of-the-art user experience, the add-on analytics module should be part of a single client-facing application. In addition, appealing visualizations with a modern look & feel will considerably improve the experience.

A single user experience typically means:

  • Single sign-on: re-use permissions and user rights of the parent application to log in users on their dashboards;
  • White-label analytics: analytics that feel native to the platform they’re embedded in. They mirror the colors, fonts, and overall look & feel;
  • Contextual enrichment: offer dashboards within the right context in the parent application. As such, users don’t need to browse through a bunch of tables and charts to find the information they need in their specific situation;

Fast deployment

Furthermore, the speed of implementation has become a crucial factor of evaluation.

To speed up development and to help developers focus on their core tasks, an ideal solution looks like the following:

  • Low-code building block: rely on a specialized solution to insert analytical power into any application. And this with the least amount of effort;
  • Plug-and-play setup: analytics building blocks easily plug into existing applications or workflows with just a few lines of code;
  • Out-of-the-box data connectors with primary data sources or applications;
  • Cloud environment: deploying analytics in the cloud makes for a scalable, up-to-date and secure solution compared to on-premise deployment;

Eventually, this leads to shorter development cycles, drastically reducing the risk of delays and setbacks. In addition, a faster deployment means faster ROI. Whether it’s generating new revenue, or faster insights in an ever-changing, competitive environment.

Self-service analytics

Of course, the implementation is just one part of the equation. On the other hand, there is the adoption by the end-users who are in search of new insights.

To keep onboarding effort and costs to a minimum, an ideal embedded analytics solution is self-serve, with self-learning capabilities:

  • Powerful & easy user interface for designers: dashboards should be easy to create. Simple drag & drop enables any business user in an organization to create & adapt reporting dashboards in minutes;
  • Actionable & intuitive interface for viewers: dashboards should be easy to consume: visually appealing, intuitive and interactive. In addition, they should automatically be consultable in multiple languages and on multiple devices;
  • Personalized insights: make use of technology that automatically personalizes dashboards based on the user’s roles & business context. Or even enable a user to create their own ad-hoc, personalized or on-the-fly dashboards;
  • Well-documented: leverage technologies that sustain onboarding for not just a handful of users, but thousands. Great online training materials, technical documentation, and regular educational webinars are equally important as ease of use.

Leveraging embedded analytics in your company

The embedded analytics market is growing. Over the coming years, we expect a lot of new things happening within the space. Businesses of all shapes and sizes benefit from faster insights & more accessible analytics:

  1. Businesses who can consume data insights from within their core applications will improve operational efficiency. Overall, they will respond faster to change.
  2. Businesses like SaaS, IoT and service providers will not only enrich their product with interactive data insights. They will also be able to generate additional revenue for their business through add-on analytics offerings.

Most industry insiders agree that world-changing events like the coronavirus outbreak have accelerated the shift towards digitalization, cloud-based applications and citizen analysts. Companies who keep up with digitalization may now wonder: “Should I consider exposing my application data to end-users? Which interest do they take in these insights?”

To answer this question, our next article will discuss exactly why companies should start exposing their data sooner than later. And moreover, how it empowers users to take data-driven decisions anywhere through next-generation embedded analytics.

Want to keep following our blog series on embedded analytics? Subscribe below and receive our bi-weekly article straight in your inbox!



Add a Comment