While much of the analytics spotlight has been on self-service, GoodData spotted a need for embedding analytics and machine learning into the operational applications people use daily.
The success of Tableau returned BI to its original promise of putting visualizations on the desktops of “everyman.” The rush to self-service grew crowded as the rest of the BI industry finally saw the path to expanding their user base beyond the foothold of elite knowledge workers or power spreadsheet users.
GoodData initially took the Cloud BI visualization path before discovering another untapped need: people working with operational applications required insights while performing their regular jobs rather than waiting after the fact and going to a separate pane of glass for the answer. That’s been the trend with new generation enterprise applications from the Oracles and SAPs of the world, who are baking analytics into their core packaged offerings. And so a couple years ago, GoodData made the pivot to becoming a platform for developing cloud-based analytic applications, and has been seeing double digit year-on-year growth since.
While GoodData’s analytic platform installed base has reached tens of thousands of companies, it’s probably the biggest analytics vendor you’ve never heard of because much of that base has come through embedding its analytic engine into third party applications. Zendesk, one of GoodData’s earliest OEM customers, provides a good example. It opted for GoodData when it faced a make vs. buy decision for adding an analytics to its service desk platform. GoodData’s analytics come through Zendesk’s UI, not another pane of glass, with the company reporting that roughly 80% of its 40,000-company installed base routinely uses the analytics on a daily basis.
The GoodData platform is a cloud-based service that includes SDKs, a REST API console which supports embedding, real-time notification, machine learning model and KPI metrics editors, a data transformation (ETL) engine, and a console for data scientists to develop predictive analytics. While GoodData has embraced machine learning like many analytics vendors, the objective is embedding the capabilities in a black box so end users consume the results as prescriptive recommendations that pop up in the dashboards they use for operational applications.
At first blush, GoodData’s embedded model appears to run counter to the trends in analytics that are currently under the spotlight. End users are getting empowered by self-service tools to move beyond being passive consumers of pre-built analytic dashboards. Meanwhile, specialists are embracing data science and machine learning with the aid of notebooks and collaborative tools to build from an increasingly vast array of libraries and algorithms that are available at their fingertips.
There is no question that traditional BI consumers are becoming more adventurous, and that data scientists are chomping at the bit to add machine learning to their LinkedIn profiles. But even as colleges and universities are turning out more data scientists, demand will not catch up to supply, at least for a long time.
But more to the point, people whose roles are not typically associated with those of “knowledge worker” can benefit just as well as their more adventurous colleagues who are exploring data. Predictive and prescriptive analytics can help the people who are processing routine maintenance work orders to reduce fraud as it can for the folks who are looking to discover where new sources of fraud are emerging. Our Trends to Watch research published earlier this year predicted that the biggest impact for machine learning would occur where it is least visible: embedded into the routine applications that people use everyday. GoodData provides a prime example of how machine learning is impacting business away from the spotlight.