Your Data Access and Transparency Journey
Your Data Access and Transparency Journey

Your Data Access and Transparency Journey

How can your organization ensure data access and transparency? Measure yourself against this model, courtesy of our partners at Slalom.

By Charles Wang, October 20, 2020

At Fivetran, we strive to promote broad-based access to data for business intelligence and data science. This is an aspiration we share with our partners at Slalom, who have developed a framework called the Modern Culture of Data. The Modern Culture of Data aims to enable organizations to shift from siloed and risk-averse practices to collaborative, experimental, and data-driven workflows.

One of the dimensions featured in this framework is Access and Transparency. This dimension contains three competencies:

  1. Data Complexity

  2. Architecture and Infrastructure

  3. Delivery and Consumption

Each of these competencies is a technological or organizational challenge that Fivetran and Slalom can help you with. Let’s discuss the opposite ends of each dimension to help you get a sense of where your company stands.

Data Complexity

In the Data Access and Transparency framework, “data complexity” refers to the sources and types of data that are produced by a company’s activities. The complexity of data increases as it becomes less structured. Structured data is supplied in a standardized format that is readily organized into tables in a relational database representing a data model. Unstructured data, by contrast, consists of formats that are not readily stored in a relational database, such as text from emails, social media, or documents, images, video, and audio.

The most basic approach to business intelligence leverages structured data from internal systems and SaaS applications. These systems and applications include enterprise resource planning (ERPs), customer relationship management (CRMs), advertising platforms, project management software, payments processing, HR benefits management, etc. You may also enrich this data with structured data from external sources such as market research and ratings publications.

At the most sophisticated level of data science, your company may use unstructured data to perform sentiment analysis, optical character recognition, and other machine learning tasks.

Architecture and Infrastructure

By “architecture and infrastructure,” we mean the tools and technologies used to extract, load, transform, and store data. In evaluating your architecture and infrastructure, you should consider the tools you have and how they are connected with each other.

A very rudimentary level of competence in architecture and infrastructure might look like the following:

  • Data is siloed and spread across a wide range of systems, with uncertain quality and provenance

  • In order to prepare data for analysis, users must manually extract, load, and transform data

  • There is no repository of data that serves as a single “source of truth”

Since the defining feature of this level of competence is manual data extraction, a company at this level of competence might have no data integration infrastructure at all. Its data team is likely to be somewhat understaffed, with each member wearing “many hats” and frequent bottlenecks caused by manual processes.

By contrast, a company with advanced architecture and infrastructure capabilities will exhibit the following:

  • Data is comprehensively integrated into one platform, with high quality and a clear provenance

  • Data is automatically extracted, transformed, and loaded

  • There is a single, well-documented “source of truth”

The data integration infrastructure at this level of competence is likely to feature an array of cloud-based tools and technologies such as:

  • A wide range of SaaS applications and operational systems serving as data sources

  • Automated data integration pipeline

  • Cloud data warehouse

  • Data lake

  • Cloud BI tool

At the organizational level, a company with advanced architecture and infrastructure will clearly differentiate between the roles of data engineering and data analysis.

Delivery and Consumption

The “delivery and consumption” dimension concerns accessibility and barriers to entry. Ultimately, the goal is to make data accessible and digestible to end-users so that data can be turned into actionable knowledge.

A company with underdeveloped delivery and consumption abilities will look like so:

  • Data is only accessible to a small number of expert users

  • Data is spread across a number of systems

  • There is very little awareness of the available data across the organization

These high barriers to entry and lack of accessibility are directly related to a company’s technical capabilities. Without the appropriate technology stack, it is impossible to centralize the data or make it accessible to non-technical users.

A company with well-developed delivery and consumption abilities, by contrast, will ensure that:

  • There are no barriers to accessing data, including for non-technical users

  • Data is accessible and easily visualized or modeled from a single system

  • All members of the company are aware of all data sources and the reports and dashboards produced from them.

Delivery and consumption are downstream of architecture and infrastructure. When you are designing and choosing the elements of your technology stack, make sure to keep the desired goals in mind.

Just One Dimension of Data Maturity

With the help of Fivetran and automated data integration, you can radically enhance your company’s data access and transparency capabilities. However, data access and transparency comprise just one dimension of data maturity. To help your company fully embody the Modern Culture of Data, consider enlisting the help of Slalom.

Start analyzing your data in minutes, not months

Launch any Fivetran connector instantly.

Adblock Detection