GoodData Insights as a Service guides users thru the analytics process

GoodData launches Insights Network that goes beyond BI and gives uses recommendations that guide them through the analytics process. I ask them about it.

By Gregory Piatetsky, @kdnuggets, Jan 27, 2015.

GoodData, a leading provider of cloud business intelligence software, today announced major upgrades to its GoodData(®) open analytics platform.

GoodData wants to let organizations to move beyond BI, a become the first "Insights-as-a-Service" provider in the market.

Gooddata made major enhancements to its Insights Engine, with the addition of Analytical Designer and Data Explorer for business user data discovery.

GoodData Dashboard
Fig 1: GoodData Dashboard, sample project.

What makes this especially interesting is the Insights Network - the knowledge accumulated over seven years of deployments, services expertise and the activity within more than 50,000 projects in GoodData's cloud-based open analytics platform. GoodData uses the Insights Network to offer recommendations within the new Analytical Designer that intuitively guide users through analysis based on data context and best practices while also helping develop their skills in data science.

(Note: A similar feature is also present in RapidMiner, which gives data analysts recommendations based on what other analysts did in similar cases).

GoodData Recommendation within Analytical Designer
Fig 2: An example of GoodData Recommendations within Analytical Designer.

I had a chance to ask them a few questions about the new release.

Gregory Piatetsky, Q1. Can you give 2-3 examples of recommendations within the new Analytical Designer?

GoodData: When a user is choosing metrics that include a connected time attribute, the Analytical Designer suggests a Period-over-Period comparison. For example, when looking at Q4 from this year, it recommends comparing it to Q4 from last year.

When picking attributes like "social channel," appropriate filters are automatically made available to users through the simple click of a button.

When picking product lines, the Analytical Designer will offer either time-over-time, granularity of time period (weeks, months, quarters, last 30 days, etc.) or a percentage of orders.

GP: Q2. How portable are analytics recommendations across domains?

GoodData: Recommendations are easily portable across domains when data sources or data types are the same.

GP: Q3. How close are you to fully automating data analysis - "drop in data, predictive model comes out" - what are the major obstacles?

GoodData: We are getting very close. GoodData helps the user understand the data that is available to him or her through the Data Explorer. The Data Explorer will simplify adding user-defined data. With this announcement, the Analytical Designer offers users suggestions for best practices in data analysis.

GP: Q4. Can you tell us, approximately, how many users does GoodData have, and how are they spread geographically and by industry ?

GoodData: Half of GoodData's business is from Technology Vendors who embed GoodData analytics within their products. These vendors serve the majority of end-customers hosted within the platform, currently about 40,000. Currently, the majority of the GoodData user-base is based in North America. However, with its upcoming European Data Center, the company is expanding business in the European Union as well.

See more on GoodData site: Welcome to the future of guided analytics: Introducing insights as a service and Insights-as-a-Service video.