Today’s best practices for embedded and self-service analytics

By | 4 minute read | October 21, 2020

The market is growing for self-service analytics

Gone are the days when line-of-business professionals could wait for analytics experts to run reports about yesterday’s results. Remember the days of combining spreadsheets, ad hoc charts and multiple versions of the truth? The need to unify data from various sources and turn it into actionable insights is more important than ever. That’s because anyone who falls behind is left behind.

The new baseline for analytics demands that data tell a story. Solutions must include embedded apps, quick self-service and real-time delivery and visuals. In every industry, users seek ready access to high quality data and analytics for evidence and confidence to carry out decisions, and the market for self-service analytics that puts business intelligence and analytics into end-users’ hands is growing exponentially.

Embedded analytics tools integrate seamlessly with applications that are used in everyday workflows. Rather than requiring users to exit a program or open standalone analytics software to find more data, integrated apps provide immediate access to analysis and visualization capabilities. Embedding keeps people engaged with the flow of accelerating insights. Advanced versions of embedded analytics offer AI, machine learning and predictive capabilities that can be truly transformative and help automate workflows, improve productivity and trigger critical actions.

Make faster data-driven decisions with integrated analytics capabilities

A report titled Next-Generation Embedded Analytics Spark Digital Transformation, authored by Doug Henschel, Constellation Research VP and Principal Analyst, paints a picture of the rise of in-context analytics, providing recommendations, best practices and a look at the horizon of embedded analytics. With several case studies that demonstrate inventive uses of embedded analytics in various industries, Henschel outlines how embedded analytics can drive business results such as increased revenue or cut costs.

“Advanced embedded analytics offer AI, machine learning and predictive capabilities that can be truly transformative and help automate workflows and trigger critical actions.” — Doug Henschel, Constellation Research VP and Principal Analyst

The evolution of embedded analytics

Performance results are often summarized with descriptive analytics that deliver snapshots of current and past business performance. But the overwhelming volume and inconsistent varieties of data threaten to overwhelm many organizations. Tools that can automate combining, cleaning and queries can help. Data visualization can clarify and explain insights more easily. And data visualizations are widely used across industries to support improvements in sales, training, management, marketing and operations.

Diagnostic analytics go even further. Business users can apply diagnostic analytics capabilities and automated data science algorithms to analytics questions and queries without the need to wait for assistance from data science experts. Embedded AI capabilities can guide the exploration of hidden relationships and drivers, and diagnostic analytics can provide answers and supporting evidence for questions like, “Why did x happen?” and “What should be done about it?” And diagnostics can pinpoint risk or flag potential regulatory compliance gaps.

Predictive analytics can reveal what may happen in the future. Self-service predictive analytics allow business users to run what-if scenarios to test hypotheses. Handy for the sandbox approach, what-ifs can help users refine a variety of parameters to try to maximize opportunities and help reduce risk. Marketing, credit scoring and crime prevention are areas where predictive analytics can be especially useful.

Innovations and disruption

Custom-designed predictive analytics, using automation and complex algorithms, can predict outcomes and recommend actions within the context of other applications, enterprise resource planning (ERP) solutions. Less commonly available off-the-shelf, such programs may be purpose-built to suit solutions use cases in a variety of industries and lines of business.

Advanced AI can forecast, execute, learn and automatically adjust to dynamically evolving scenarios, using continuous internal feedback to self-regulate and learn, adjust and execute based on results developed by the program itself. Once the wave of the future, these applications going mainstream. Innovative analytics programs can help enterprises disrupt markets and create new ones like programs that match parties in uneven supply-and-demand markets, like ride-hailing apps.

Identify your BI capability goals and deployment parameters

The Constellation Research report recommends a list of considerations to help identify organizational goals for building business intelligence (BI) capability, from defining the pain points and opportunities, to proposed user groups and skill levels, current BI capabilities, and deployment parameters. The report also details key questions to consider when reviewing potential analytics vendors.

Considerations include:

  • Flexibility of architecture and deployment
  • Data management, data science, customization and extensibility provisions
  • Security requirements 

Built for the future of embedded analytics

IBM® Cognos® Analytics is an example of an AI-infused solution that offers self-service descriptive analytics, forecasting, and diagnostics. Cognos is built to be embedded into your applications so that users can build reports, dashboards and highly engaging, interactive visualizations derived from virtually any data source.

With support for REST and JavaScript APIs, it is faster and easier to develop applications that embed, customize and automate Cognos Analytics functionality. For example, the dashboard API is primarily geared to embed the interactive widgets of a dashboard in a web page or application with full control over layout and styling.

Users can access most data sources directly from a wide variety of applications including transactional OLTP databases, data marts and warehouses, big data clusters, columnar, in-memory or OLAP sources. This is true even for extremely complex data sources with multiple databases, tables and complex joins. The data can live on-premises, on cloud or in hybrid or multi-cloud environments.

Cognos supports your existing security model for users, groups and roles, supporting single sign-on with almost any security provider. Next-generation RESTful APIs enable application development for extending, leveraging, and automating functionality.

Learn more about how to drive more confident decisions at Cognos Analytics or read the full report from Constellation Research.