Our Data Methodology: Infrastructure Architecture

April 28, 2020

How to get the right data to the right person to make the right decision at the right time.

This is part one of a two-part series about our approach to establishing a strong foundation for advanced analytics.  

At its most powerful, data is a business-driving asset. But the mere collection of data doesn’t add value. The practice and service of data do. And they both require a supportive infrastructure—infrastructure that gets the right data to the right person to make the right decision at the right time. Of course, this is easier said than done, but we’ve developed a methodology to help you get there.

Infrastructure architecture

When we use the phrase “infrastructure architecture” with our clients, we’re referring to the people, architecture, and time required to support advanced analytics. We know: Investing in this foundational effort is about as sexy as investing in the plumbing of your home: It’s not always fun, but the alternative is much worse. If you invest in an infrastructure that supports the FAIR (findable, accessible, interoperable, and reusable) principles of data, the return on your investment will be a quantifiable asset. And that, after all, is why we’re all here.

FAIR data empowers you to not just report data, but to make data-driven business decisions, whether that means creating new products and services or making more informed hiring decisions. To that end, we recommend a layered approach to data architecture, where each layer of the architecture forms a specific abstraction that satisfies a catalog of business requests. 

Here’s what it looks like

SingleStone Data Methodology
Click to enlarge

The practice of data, the wrangling of oodles of data into a structure that enables intelligence usage, occurs in layers 0, 1, and 2. The service of data, constructing a user interface that enables the consumption of data for decision making, occurs in layers 3 and 4. Over time, this layered approach to architecture reduces redundancy, decreases dark data, and increases your return on investment earlier in the data lifecycle. So, how do you get started?

  1. Identify the data stewards in your organization. The most evolved data-centric companies identify data stewards to oversee the architecture of infrastructure and engagement (you can read more about engagement architecture in part two of this series). These stewards are experts in both the business and your customers. They understand what data you’re collecting and why you’re collecting it. They also understand the data’s limitations. 
  2. Identify who’s accessing the data and why. This will help you define the business requirements that will dictate your architecture design. At the consumption layer, executive decision-makers are consuming reports with historical, current, and forecasted data. That’s why it’s sometimes also called the “Executive Layer.” The biggest challenge in this layer is designing an easy-to-use interface that informs the consumer without influencing her. I often see visualizations and think, “It’s pretty. And it’s pretty wrong.”
  3. Don’t ignore your dark data. Most companies are making decisions based on only 10% of all of the data they’re capturing. The other 90% is called dark because it has never seen the light of day—it hasn’t been touched. Progressive practitioners know that this dark data can be a goldmine of insight. After all, if you’re making decisions based on a fraction of the data, how can you be sure that you’re making the right decision? 
  4. Take stock. Do you have appropriate and abundant resources to build and operate these layers? Many organizations begin this work in one vertical. Taking this topical approach gives you the biggest bang for your initial buck. But it’s only when you do this at an enterprise level that you’re truly able to innovate. 

Infrastructure architecture is the foundational layer of how we approach advanced analytics. Once these principles are understood and addressed, we move on to engagement architecture.

Sign up to hear more below, and follow us on LinkedIn.

Read Part Two Download our Data Methodology Visual

Sign Up to Hear More

Get notified when more posts like “Our Data Methodology: Infrastructure Architecture” are available.

  • This field is for validation purposes and should be left unchanged.

Vida Williams

Vida Williams, our Advanced Analytics Solutions Lead, began her career in technology as a tech writer. Today, she uses her experience as a data scientist to drive social and economic change. Vida is passionate about data inclusivity and serves as the first-ever “Innovator in Residence” at VCU’s da Vinci Center.

Leave a Reply

Your email address will not be published. Required fields are marked *