Organizations are creating an enormous amount of data, from a growing number of sources. The task of managing it all falls on IT, but determining success with any analytics project starts with business users.
Information Management spoke with Shikhar Agarwal, founding engineer at ThoughtSpot, for her observations on what attendees at the recent Strata & Hadoop World conference in San Jose, CA, aare trying to do with their business intelligence and analytics efforts, and the questions they should be asking themselves to determine if they’re on the right track.
Information Management: What are the most common themes that you heard from conference attendees and how do those themes align with what you expected?
Shikhar Agarwal: With more data comes more responsibility. As companies are increasingly equipped to generate and store massive sets of data, IT must now take on the nearly impossible task of managing, provisioning, governing, preparing, and analyzing growing data volumes with increasingly nuanced expectations from line of business users. We keep hearing, at Strata and other conferences, that as business intelligence and analytics expand beyond the purview of traditional IT roles, tech investments are be tied to and measured by their business impact. Some questions that organizations should ask:
Are line-of-business users able to uncover insights on their own and make decisions quickly?
Is the information accurate and timely?
Do they improve campaigns, programs, products and policies, thus driving the business’s bottom line in a positive way?
There’s a dirty secret in the world of BI: most BI or analytics projects are plateauing at a dismal 25% adoption rate. That’s hardly enough to have meaningful impact cross-functionally.
Everyone — from line of business users who need access to insights; to the data science teams who, strapped for time and resources , need to focus on more strategic data initiatives — are clamoring for technology that puts analytics and insights into the right hands, at the right time.
IM: What are the most common challenges that attendees are facing with regard to data management and data analytics?
SA: Data prep came up a lot. Many new BI products that have launched over the past few years fulfill one or more of the self-service data prep steps: data discovery and profiling; catalog and metadata; data structuring and modeling; data transformation; data curation; enrichment and collaboration. Despite this, many BI professionals who operate in complex data environments still spend up to 90% of their time preparing data for analysis. As data volumes balloon, companies know they have to adopt new technologies that automate a good deal of this legwork. The alternative — to continue to lock precious data science resources into routine data preparation — simply isn’t feasible.