Data Silos Are The Death of Analytics. Here’s The Fix.

This statement in a recent Information Management article caught my eye: “…analytics and new insights can not only differentiate you from the competition, but in some cases, crush the competition.”

We’ve seen customers grow increasingly interested in using analytics as a competitive weapon in a disrupt-or-be-disrupted marketplace. But for all the fancy tools you can invest in, your analytics will only be as good as the data behind it.

The bottom line is this: if your analytics initiatives aren’t supported by an active data management architecture that’s applied across all your projects, you won’t be able to keep up with your competitors that have a data strategy.

It’s all about delivering actionable insights faster.

Many IT leaders understand this, but analytics initiatives aren’t funded as top-to-bottom data-led programs. Instead, IT is given specific projects with specific goals, and tight time/budget constraints. The temptation with each project is to deliver on the business goal in the fastest, most efficient way possible. Project after project, there’s little or no thought about connecting these tools and data sets with previous projects or future ones.

When Informatica interviewed 150 architects on their challenges around analytics and data governance, we saw the problem vividly. A whopping 60 percent said they knew exactly the trouble their company would get into without an overall vision for data architecture—and yet they didn’t feel empowered, or budgeted, to do much about it.

A Repeatable Data Management Strategy for Analytics

data ebook1
Be decision ready!

The answer, we think, is to envision a comprehensive plan for how your company handles data. Then, design each project to work with the plan, moving you closer to your coherent strategy. In short, stay out of the data silo business.

With that challenge in mind, we’ve written an eBook designed to help organizations define their data challenges, from status quo to competitive threats to road forward. “ What It Takes to Deliver Advanced Analytics ” is designed to help you approach analytics in a comprehensive way that will allow you to think big, and create a framework that can guide your decisions on specific efforts along the way.


data ebook2
The six essentials of great data

In a fast-paced 20 pages, the book discusses the six essentials for a data-driven business:

  • Repeatability
  • Abstraction
  • Alignment
  • Collaboration
  • Automation
  • Governance


We discuss the importance of each, and most importantly, provide action items to let you tackle their challenges. You can download the full book right now, but here’s a preview of our approach: a look at those first two topics.

Repeatability: Once Isn’t Enough

In big data analytics, 50 percent to 80 percent of the work is in the data wrangling—identifying the data you need and pulling it into a clean, coherent data set. The actual analytics work might be only 10 percent of the project cost. The key to minimizing repetitive work is finding a way to easily reuse your logic on the next data set, rather than starting from square one each time.

The eBook offers four action points:

  1. Start by documenting your current processes for crucial steps that come up every time. Things like data cleansing and data integrations.
  2. Analyze the processes to find out how long they take and whether or not your team follows the same steps every time.
  3. Identify opportunities to standardize these steps and create a common roadmap and logic for teams to follow.
  4. Once you know these details, look for tools that can help you standardize these steps in an efficient and scalable way (more on this in “essential” #5).

InformationWeek’s “ Analytics and BI Survey 2015 ” found that the top three barriers to successful information management are accessing, integrating, and cleansing data. Remove these obstacles, and lower the time you spend wrestling with your data, to let yourself extract the insights your business needs much faster.

Abstraction: Rise Above Ground-Level

It’s vital to pursue a data management architecture that works across any type of data, BI tool, or storage technology. If the move to add Hadoop or NoSQL demands entirely different tools to manage the data, you’re at risk of creating another silo.

Your data management tools should work across any type of data, platform, engine, or use case. When you’ve got different tools for your traditional data warehouse versus your cloud setup, and therefore different skill sets to hire for, train for, and maintain, you’re looking at a real mess. You want one environment that abstracts you from that, so you can do prep work regardless of what the underlying technology is.

The eBook offers two critical action points:

  1. Choose data management tools that abstract your data from the underlying storage technology.
  2. Make sure your integrations and cleansing processes take place in the data management layer rather than in applications.

The technology behind data storage for analytics is evolving very quickly. Architect your data management platform to abstract you from that change. The flexibility will keep your teams and processes decision ready.

Building a Better Analytics Infrastructure

We’ve created this eBook to help architects who see a bigger picture create success in a world where the project is always the imperative. (There’s also an in-depth workbook for tackling those projects—the subject of my next post.)  It’s not too late to change your organization’s overall management of data, and to do so in a way that lets you continue to satisfy the “solve today’s problem” approach to analytics projects.

Download What It Takes to Deliver Advanced Analytics: The Six Essentials for a Decision-Ready Business today, and check this space next week for the next phase in your organization’s evolution.