centrally driven

In Thomas Davenport’s “Competing on Analytics: The New Science of Winning”, one of the first chapters of the book defines the common attributes of analytically-driven organizations. Davenport discusses that one of the critical aspects to success in analytics revolves around taking an enterprise-level approach to managing data and analytics. He then quotes Harrah’s Entertainment’s management approach to enterprise data and analytics, calling the approach “centrally driven, broadly distributed”.

Focusing on “Centrally Driven”

In this article – part 1 of a two part post – we will discuss the first half of that quote. Having data and analytics be “centrally driven” is of the utmost importance to success and sustainability of your analytics programs.

Inconsistencies Abound

If you are an executive of any industry, I know for certain you have experienced the following scenario: you ask the same data and analytics question to different departments and you get several different responses.

If you are in the banking industry, ask Marketing, Lending, IT, Finance, Commercial, and Retail Operations how many customers/members you have.

How many different answers do you think you will get?

In most organizations, each department would independently retrieve the information that you asked of them. Marketing might go to their MCIF, IT might go to the core, Finance might look at the GL application – rarely will each department go to a centralized, single source of truth to retrieve the “correct” answer. Since each application is managed separately with different data contained within, there is no way to ensure consistency and accuracy without a central data warehouse or repository.

Without a centralized, enterprise-level platform of data and analytics, it is nearly impossible to ensure consistent, accurate reporting.

Spreadsheet Errors

I love Excel as much as the next guy. Quite a few organizations across nearly every industry live and die by their Excel spreadsheets (investment banking first year associates know this all too well). But how many Excel spreadsheets are 100% correct and error-free? Some studies estimate that nearly 9 in 10 Excel files have errors!

Especially in the banking industry, I see organizations littered with complicated spreadsheets. Analysts will extract Excel files from different data sources and then hope they’ve defined their VLOOKUPs accurately to be able to consolidate the data sets. I’ve personally witnessed high-level, executive reports with wildly inaccurate formulas that effectively rendered the spreadsheets useless. This is, sadly, not an anomaly.

How do you resolve this issue? By taking a centrally driven approach to managing data, users will not be required to consolidate data from various sources or manually define complicated, error-prone formulas. Data will be consolidated and validated reducing or eliminating the vast majority of these overly-complex Excel spreadsheets. Data warehouses combined with BI and data visualization tools (like Tableau, InformationBuilders, Qlik, and many others) provide an enterprise-level, industry-leading platform for data and analytics.

Definitions and Data Governance

Whenever I get the opportunity to speak with a credit union, I eventually always ask the same question, “what is your definition of a member? Does everyone have the same definition as you?” This seems similar to the inconsistencies idea brought up at the beginning of the article, but even with the same data source, there is no assurances that everyone will define a key business term the same.

Marketing might only focus on members that have not been placed on a “do not contact” list. What about the same individual (i.e. unique SSN) that has multiple accounts and member/customer numbers? How are they counted for aggregation purposes? These subtleties are critical to a strong data and analytics program. A cross-departmental team must agree on key definitions that are used throughout the organization. This centrally driven approach to defining key business terms helps ensure accountability and consistency.

Wrapping up

A centrally driven approach to data and analytics ensures consistently accurate reporting with key business definitions universally understood by everyone in the organization.

In our second part of this post, we talk about why having a “broadly distributed” data and analytics solution is vital.

Check back soon or subscribe to our blog for updates!


1 reply

Trackbacks & Pingbacks

  1. […] our previous post, we introduced you to the data and analytics management methodology, “centrally driven, broadly distributed”. Having discussed the reasons why a “centrally driven” approach to data and analytics is […]

Comments are closed.