What Exactly Is a Data Lake?

A data lake is a centralized storage repository that can hold a large amount of both structured and unstructured data at any scale. Unlike a data warehouse, which stores data in files and folders, a data lake uses a flat architecture to store data. In this architecture, each element is assigned a unique identifier and tagged with additional metadata tags. All data is loaded from source systems and no data is turned away.

 

Do I Need a Data Lake?

When properly managed and applied, data lakes are recognized as helping organizations successfully generate business value from their data. They allow organizations to perform new types of analytics like machine learning, visualizations, and big data processing. The types of data you can analyze also increase to include social media feeds, click-streams, and internet connected devices.

 

Data Lake Use Cases

 

Store and Analyze More Data

Because it stores all data, not just data that is immediately needed, a data lake allows you to go back in time to do historical analysis. Also, the structure of the schema for your data is not defined when data is stored. Data lakes are designed to store non-traditional types of data such as sensor data, social network activity, web server logs, texts and images. These types of data inputs are kept in their raw form and only transformed when they are ready to be used. The approach is known as “Schema on Read” instead of the data warehouse approach which is known as “Schema on Write.”

Deep Dive into Data Sets

The majority of employees at any given organization that manages data will not require very detailed analysis of the data they store. However, there are other individuals in the organization who are required to conduct a more deep dive analysis into their data sets. For these employees, that data lake approach can be very beneficial because of the amount of data stored. It allows them to mash up many different types of data and surface entirely new questions to be answered. Many times, these individuals are given the title Data Scientist and they take advantage of advanced analytical tools and capabilities like statistical analysis and predictive modeling.

The Desire to Have a More Flexible Data Approach

One of the biggest downfalls to the data warehouse approach is the time it takes to alter its structure. Even the best data warehouse design will struggle to adapt to change because of the complexity of the data loading process. Considerable developer resource time is also needed to make analysis and reporting easy for the consumer of the data warehouse.

However, with the data lake approach, all data is stored in its raw form and is accessible on demand for users. This allows individuals to go beyond the structure of the warehouse and uncover meaning behind data in new ways.

Gaining Faster Insights with Data Lakes

With data lakes, users are able to get results and insights from their data faster. This is because data lakes contain all data and data types and provide access to data before it has been transformed, cleansed and structured. The downside, however, is that the time necessary for users to explore raw data as they see fit can be substantial.

 

Key Capabilities of a Data Lake

 

Analytics

Data Lakes allow various user-groups in your organization like data scientists, data developers, and business analysts to access data with their choice of analytic tools and frameworks. This includes open source frameworks such as Apache Hadoop, Presto, and Apache Spark, and commercial offerings from data warehouse and business intelligence vendors. Data Lakes allow you to run analytics without the need to move your data to a separate analytics system.

Securely Store, and Catalog Data

Data Lakes allow you to store relational data—operational databases, and data from line of business applications, and non-relational data—mobile apps, IoT devices, and social media. They also give you the ability to understand what data is in the lake through crawling, cataloging, and indexing of data. Finally, data must be secured to ensure your data assets are protected.

Data Mobility

You can import any amount of data into a data lake and the data can even be ingested in real time. This approach allows data to be collected from many sources, and moved in its original form. Through this process, you can scale to any size data and also save time creating data structures, schema, and transformations.

Machine Learning

Organizations who use data lakes can generate numerous different insights including reporting on historical data. They can also create a form of machine learning where models are designed to forecast likely outcomes and suggest prescribed actions to achieve a desired result.

 

The Value of a Data Lake

The ability to tap into more data pulled from more sources and in less time empowers users to collaborate and analyze data in different ways, which leads to faster decision making. Some of the areas data lakes can add value include:

Improved Customer Interactions

Data lakes allow you to improve your customer interactions because it provides you with more information about your audience that you can analyze to better understand their needs and sentiment. For example, you can combine and analyze customer data from a CRM platform with social media analytics as well as a marketing platform to understand and profile profitable customers, the cause for churn, and the optimal digital marketing strategy.

Improve R&D Innovation Choices

With a data lake, R&D teams can better test their hypothesis, refine their assumptions, and assess results. An example could be understanding data around the desires for certain features in products and then understanding how they can best be implemented.

Increase Operational Efficiencies

With the rise of the Internet of Things (IoT), more and more data is being introduced into all types of industries, with much of this data being streamed in real-time from internet connected devices. A data lake makes it much easier to store, and run analytics on machine-generated IoT data to uncover ways to reduce operational costs, and increase quality.

 

Is the Future of Data Lakes in the Cloud?

For many, the term data lake has become synonymous with big data technologies like Hadoop. This is because it scales and adapts well to very large volumes of data and can also handle data in any structure. However, it has become apparent that Data Lakes are an ideal workload to be deployed in the cloud, because the cloud provides performance, reliability, availability and scalability, as well as a diverse set of analytic engines, and massive economies of scale.

Some of the primary reasons customers see the cloud as an advantage for Data Lakes are better security, better availability, faster time to deployment, more frequent feature/functionality updates, more elasticity, more geographic coverage, and costs linked to actual utilization.