Driving success in analytics programs

Data Analytics August 27 Resolve Tech Solutions

Enterprise analytics programs, when implemented successfully, can transform a company’s business. On the one hand, employees are freed from the laborious hours of producing & digesting spreadsheets; on the other, executive management becomes more confident with their decision making. While the efficiency brought in by an analytics program may be easily quantifiable, measuring the value of effective decision-making is downright difficult, if not impossible.

With much riding on the success of analytics programs, why do many of them fail?

  • Gartner predicts that through 2022, only 20% of analytic insights will deliver business outcomes. [1]
  • According to VentureBeat, only 13% of data science projects make it to production. [2]
  • In an executive survey by NewVantage Partners, 77% of the respondents said that business adoption of big data and AI initiatives continues to represent a challenge for their organizations. [3]

In spite of this risk, organizations continue to invest in analytics programs since the returns are incredibly lucrative. Given this interest in analytics programs, following are some success factors that can help enterprises drive home their analytics programs.

Don’t boil the ocean

Given the advancement in storage and network technologies, it is alluring for the CIO to set up a data lake and start capturing all the data available: customer profiles, product specifications, ambient parameters from IoT sensors, and data from social media sites, to name a few. However, it is important to note that a comprehensive data-strategy does not imply capturing all data comprehensively. Instead, the organization would be better served by an agile approach that demonstrate value to business users – by identifying & transforming pilot business areas that are suffering from lack of insights.

Involve executive leadership

Executives often lack a solid understanding of analytics concepts, and the difference between traditional analytics (such as BI & Reporting) and advanced analytics (tools such as Machine Learning & Big Data)[4]. Naturally, they may not fully grasp the business value of their analytics programs, or worse, may identify use cases that are inappropriate for the analytics team’s skillset & orientation. The executive team needs to attend structured primers and workshops that help them understand the fundamentals of data sciences & analytics.  As an ongoing effort, the analytics program should identify champions, employees who can speak in the language of the business and evangelize the cause of the program among executive leadership.

Empower users

No matter how advanced the technology and data science concept behind the analytics program, it needs to be designed with the business user in mind. While business users may be enthralled by the idea of not having to create & populate cumbersome spreadsheets anymore, they will be disappointed quite easily if they have to learn esoteric table names and master SQL syntax. Any analytics implementation, therefore, must focus on ease of use – it should include data terms & nomenclature that the user is already familiar with and be delivered via a simple self-service interface.

Select technology carefully

There is an abundance of analytics tools available covering functionalities for ETL, data storage, business intelligence, and reporting & visualization. It is important to recognize that most of these tools, even if they are sold by the same vendor, could actually be bespoke applications allowing them to be evaluated independently. The enterprise therefore should make use of this opportunity to evaluate each tool individually for the fit for their analytics program.

Just a couple years ago, Business Intelligence dominated the analytic landscape, encompassing and integrating with a variety of technologies[5]. Today’s interpretation of enterprise analytics, however, encompasses a wide range of faculties geared toward helping organizations take better & faster data-driven decisions. Coupled with the variety of technology options available, it is incorrect to expect a one-size-fits-all program structure. The above guidelines, therefore, offer a starting point for designing a successful analytics program.


References

  1. Andrew White, Our Top Data and Analytics Predicts for 2019, https://blogs.gartner.com/andrew_white/2019/01/03/our-top-data-and-analytics-predicts-for-2019/
  2. VentureBeat, Why do 87% of data science projects never make it into production? July 2019 https://venturebeat.com/2019/07/19/why-do-87-of-data-science-projects-never-make-it-into-production/
  3. NewVantage Partners LLC, Big Data and AI Executive Survey 2019, https://newvantage.com/wp-content/uploads/2018/12/Big-Data-Executive-Survey-2019-Findings.pdf
  4. McKinsey & Company, Ten red flags signaling your analytics program will fail, May 2018https://www.mckinsey.com/business-functions/mckinsey-analytics/our-insights/ten-red-flags-signaling-your-analytics-program-will-fail
  5. John Boyer, et al., 5 Keys to Business Analytics Program Success, © IBM Corporation