Gartner Symposium/ITxpo: Getting IT Out of the Analytics Equation
Data Discovery Is Key to Adoption of Self-Service Business Intelligence
We’re finally at the Gartner Symposium/ITxpo in Orlando and it is shaping up to be a great event with close to 10,000 attendees. The team behind entellitrak will be participating in sessions across the domains of BPM, Dynamic Case Management, Business Intelligence/Analytics and Healthcare, as well as several that focus on the latest technology trends.
We also debuted our new Natural Language Analytics module at the Symposium on Monday. (We are in Booth 903, so please stop by for a test drive if you are able.)
As we start this exciting week, I wanted to shed some light on the rapidly growing field of business analytics and why there is a need to make analytics more accessible to the business user.
Google the word “analytics” and it comes back with a very succinct (and precise) explanation: “The systematic computational analysis of data or statistics.” Wikipedia defines analytics as the “discovery and communication of meaningful patterns in data.” I think that the word “systematic” in the first definition relates to the ability to drill through the data in an intuitive manner, analyze it and get to the relevant information. Wikipedia’s definition not only highlights the importance of data discovery but also the need to communicate the patterns found in the data with others, and to be able to collaborate on it to attain effective business outcomes.
With the volume of data across industries growing exponentially - especially in the government sector - there has been a growing demand amongst decision makers to be able to analyze and understand the power of data they have in their systems in real-time.
Traditional Business Intelligence (BI) systems have been performing this function for a while now. However, as customers become accustomed to a technology their needs grow and they want more from their existing systems. BI platforms tend to be very IT-driven and reports-centric. They generally have long implementation cycles and, by the time you build the report, the original data has become stale from an analysis and decision-making perspective. BI tools are challenging in that most new reporting requirements require business users to go through an approval process and waiting period for IT to deliver the report while in most cases, business would like the report right away. Data discovery capability in traditional BI systems is very limited and is mostly based on pre-defined paths, which in turn restricts users from getting to the data that they need for their analysis and reporting purposes.
Admittedly, to handle complicated scenarios it will most often be necessary to involve IT as well as business analysts. Oftentimes however the needs of end-users are quite straight forward: They just need a quick answer for a question that may have come up in a meeting. Listed below are some of the reasons I believe analytics solutions should be more flexible, accessible, and business user friendly so the dependence on IT can be minimized:
- Dependence on IT restricts business users from fully exploiting the power of their data.
- Data discovery process is not straightforward and not very end-user friendly. A significant learning curve is involved for the business user.
- Even simple use cases require someone to write SQL scripts to pull the data.
- Users may not have direct or ready access to the IT team.
- Requests may be subject to internal politics, conflict of resources, time constraints, etc.
- Depending on the architecture, the data may not be fresh.
In a report published by Gartner in April 2013, “Market Trends: The Collision of Data Discovery and Business Intelligence Will Cause Destruction,” Gartner Analysts Dan Sommer and John-David Lovelock stress that visualization-based data discovery will become a major driver for BI adoption among business users, moving mainstream business users from descriptive (what happened) to more diagnostic (why it happened). According to the report, the data discovery segment of the BI market is going to grow much more rapidly.
So what needs to happen to minimize the dependence on IT when it comes to making analytics more useful for business users? Based on my understanding of the analytics products, below are some of the areas I believe can help business users get the most out of their data.
Build flexibility into analytics tools to do ad-hoc and complex data discovery
Natural language based search
Even if IT provides all the reports that the business user needs there can always be a scenario that is not covered in the canned reports. That is where the natural language based search and data discovery can come very handy. Let’s explain it using a simple use case.
A manager has to provide feedback to his/her supervisor on all the sales for a particular item that happened over the last two weeks outside of U.S. In a traditional BI environment, data does not get refreshed as quickly. Even if it does get refreshed, a view of international sales may not be built into any of the reports. With natural language search a manager can just type in the query such as “Find all the sales for the previous 2 weeks by country not USA.” As long as the data stores country information against each of the sales, the system should be able to generate the most appropriate view (pie chart, column chart, etc.) to show the data. A tool should also allow the user to disambiguate the data if the output does not reflect the exact scenario requested. Also, the user should be able to then save this report on the fly and add it to his/her dashboard.
Allow business users to filter the data/reports across multiple dimensions (attributes), measures and date/time (by days, weeks, months, quarters, years). This allows users to see data patterns and behavior in several different modes. For example, imagine you are looking at a chart that shows all the complaints filed over the last five years broken down by year. You may want to filter this data to see the complaints originated in just the northeast region for last five years. A user should be able to create a modified view of the report by picking the right filters, without having to go to a business analyst or other IT folks for help.
The ability to drill across dimensions, measures and date/time is key for advanced data discovery. This allows the user to get deeper into a specific data point and see how it ties in with other attributes. Let’s say that for a really high number of complaints filed in a particular month, you would like to drill down and see which regions the complaints filed from or who filed those complaints. Or, maybe you want to drill down and see the filing dates for each of those complaints in that particular month.
Allow business users to create reports on the fly
As discussed above when business users have multiple ways to dive into their data and create scenarios on an ad-hoc basis, the next thing they would want would be to save their results in specific reporting formats to present to their superiors or for team discussions. If business users can just save the output of their data discovery as a report on-the-fly (instead of going to IT to create a report based on a use case) and embed that in a dashboard, such functionality will greatly reduce the dependence on IT and save business analysts' time.
Although not directly tied to IT dependence, the ability to collaborate with others on reports is a key motivator for business users to play and explore their data, as well as discover patterns that eventually help them to make effective business decisions. With the advent of social media capabilities, some analytics tools now allow users to do real-time collaboration by entering comments, provide feedback on the reports and on specific data points (especially those that may be red flags) and sharing them with others.
With the evolution of more advanced technologies and significant improvement in performances, many smaller vendors are coming out with analytics tools, which allow analysis to be done directly in the data source and in-memory. Many of these tools allow faster extraction of data even if the dataset needs to be separated from the primary data source by passing the need for expensive, time-consuming ETL processes and allowing users to play against their entire data rather than a subset of it. Some of these technologies are still new and need to mature further but the trend is clearly visible. As per Dan and John in the above mentioned Gartner report, data discovery will become a key tenet of analytics putting business users in the driver’s seat.
We can continue this discussion across several impact areas and how vendors can make their analytics offering much more appealing to the business users but for lack of space I will limit my discussion at this point.
Feel free to check out one of my previous blog posts Analytics Insights for the Business Process that discusses the emerging models of analytics.
The new Natural Language Analytics module will be offered to the customers as an add-on module along with the BPM and Case Management platform entellitrak. This analytics offering from MicroPact also aligns closely with its Data-First™ approach to dynamic case management. You can download this free CITO Research report titled “How Data-First Development Overcomes Barriers to Creating Effective Applications.”
You can download the analytics datasheet by going to this page on MicroPact’s website. For any additional information on Natural Language Analytics offering please contact the MicroPact sales team at email@example.com.
- Answering Your Questions about Partner Sales Enablement February 5, 2020
- Introducing Chris Flores, Vice President of Global Alliances January 15, 2020
- 6 Tips for Stress-Free Software Adoption December 11, 2019