Application teams looking to embed analytics can choose from dozens of embedded analytics tools. “But not all analytics platforms were built to be embedded,” writes Gartner in a new report, 5 Best Practices for Choosing an Embedded Analytics Platform Provider. The recommendations in the report are based on calls with a dozen enterprise application providers offering embedded analytics.
What guidelines should application teams follow when selecting an analytics platform? Here are the main points from the Gartner report, along with supporting evidence from Logi’s own 2018 State of Embedded Analytics Report (based on our survey of 500+ members of application teams).
#1 Match the embedded analytics capabilities to user needs and skills
Understanding who will be using your application, along with what questions they are trying to answer, is paramount to finding the right analytics platform. By tailoring the analytics experience to each user’s skills and needs—from casual users to data scientists—you will increase user adoption.
“To minimize the need for training and to increase usability and adoption, look at ways to match and adapt the analytics for the user personas,” writes Gartner. “Don’t make users spend precious time adapting and learning the embedded analytics tools.”
The 2018 State of Embedded Analytics survey confirms that this practice pays dividends. Ninety-one percent of application teams said embedded analytics has helped them increase end-user adoption.
#2 Match the embedded analytics tool to the rest of your application
A seamless user experience is crucial for analytics success. That’s why embedded analytics tools are superior to standalone analytics solutions, because they deliver the look and feel of a single application rather than two different products. Gartner advises choosing an embedded analytics tool that allows you to rebrand (or white-label) the user interface so it looks like part of your broader application.
According to Gartner, “This seamless approach means that users don’t even know they are using a multiproduct application. It allows users to take immediate action from within the application, without shifting context.”
No matter what analytics vendor, platform, or individual components you use to power the dashboards and reports in your application, your customers should always know they’re seeing your brand—not the brand of your analytics platform. White-labeling analytics means matching the colors, fonts, and design elements to your application. But it’s also important to ensure white-labeled analytics interact with your application’s other features. The ability for users to kick off a new workflow, update the database, and share findings with colleagues all add up to a feeling that everything is part of the application’s native features, and not some bolted-on analytics from a third party.
#3 Assess the capabilities by looking beyond visualizations
When it comes to embedded analytics platform providers, Gartner cautions against focusing too much on what you see on the front end. They suggest choosing an embedded analytics tool that not only looks great but also works with your application on the back end. A solution should be able to scale and grow with you, work with your security model and data architecture, and offer compatible deployment options.
“The easiest way to understand the limitations of the embedded analytics platform is to request a proof of concept rather than a demonstration,” says Gartner. “Use your own data, not sample data, to see how accurately and quickly the tool can get embedded analytics into your application UI with usable, interactive data.”
#4 Add machine learning to augment decision making
According to Gartner, machine learning (also known as augmented analytics) can increase the value of your embedded analytics in many areas, including data prep, natural language interfaces, automatic outlier detection, recommendations, and causality and significance detection. In a nutshell, all of these features help speed up user insights and reduce decision bias.
“Embedded machine-learning-based analytics can make your application more pervasive and make it more useful to a broader set of users,” notes the Gartner report.
Logi’s survey confirms that these types of sophisticated analytics features bring big business benefits. In fact, compared to application teams that offer only basic analytics features, companies that deliver sophisticated analytics are better able to differentiate their products from competitors and reduce their customer attrition rates.
#5 Search for a provider that focuses exclusively on embedded analytics
Gartner advises that evaluating the analytics provider itself is just as important as the capabilities of the analytics platform. This is partially because you want a trusted partner, but also because you will be working with your analytics provider for the long term—and it’s a major effort to switch to another provider later.
“Providers that focus just on selling embedded analytics (or a significant portion of their revenue is from embedded) generally understand what is needed to ensure your success. They have dedicated engineering and service support to help you integrate and embed the analytics into your application,” notes Gartner.
Our survey shows analytics solutions are taking longer to deploy, in part because of their growing complexity. Eighteen percent of our survey respondents say it took them longer to get to market than expected, and more than 25 percent say the people and resources required were greater than expected. Considering the fact that application teams spend significant time and resources to deploy analytics solutions, it’s increasingly important to choose a vendor that’s completely focused on embedded analytics.