Do you Need a Warehouse, a Lake or a Mesh?

Posted by Tori Williams | 10-Dec-2021 15:16:10

To remain customer centric, enterprises must understand ever expanding volumes of data and it has become business critical to utilise a data architecture that best reflects the needs of employees and customers. Two approaches have traditionally dominated: the data warehouse and the data lake. Recently, a third contender has arrived: the data mesh.


Since the 1990s, organisations have gathered and processed business information in data warehouses. Gartner defines a data warehouse as

“A storage architecture designed to hold data extracted from transaction systems, operational data stores and external sources. The warehouse then combines that data in an aggregate, summary form suitable for enterprise-wide data analysis and reporting for predefined business needs.”

A key benefit of a data warehouse is that it stores highly structured, refined and pre-processed data, and this can aid business analysis and decision making for specific use cases. So, it is a useful option for businesses or departments who use data predominantly for querying and reporting. 

But as enterprises seek to process data at scale and in real-time to better serve their customers, traditional data warehouses can struggle to provide the agility and responsiveness modern enterprises require. Businesses also find that upgrading and reconfiguring on-premises data warehouses can prove to be expensive and time-consuming.


An alternative is a data lake. Data lakes are often quicker and more cost effective to set up and operate, as they move away from the rigid structure of the data warehouse. As a data lake stores both structured and unstructured data, they can also provide fertile ground for artificial intelligence and machine learning solutions. 

Data lakes based in the cloud also enable greater scalability, flexibility and responsiveness compared to traditional data warehouses - features which are well suited to the needs of modern enterprises, looking to analyse streams of data from e-commerce, entertainment and social media sources in real-time. 

Although, this constant flow of data can create an ever expanding data lake in which identifying insightful information among tonnes of raw data can prove challenging. To have a masterful handle on a data lake, organisations require highly skilled data scientists and analysts. But this can also mean useful insights are hard to reach for the rest of the business and siloes can emerge. 


A new architectural concept, the data mesh, promises to fuel big data innovations while addressing the governance challenges data lakes often encountered. This is achieved by taking a decentralised approach to data ownership. 

This architectural style has been gaining traction since Netflix implemented a data mesh architecture, which helped to manage cost and optimise performance. However, as a new and quickly evolving approach in the market, there is still much to learn, test and trial when it comes to choosing and implementing a data mesh to work for the needs of individual businesses. 

To learn more about this new approach, and hear how leading enterprises are choosing and implementing their data strategies to reflect the changing needs of consumers, join our lunchtime session in partnership with mParticle from 12:00 - 12:45 on Thursday 20 January. 

If you’re looking to find out more about how to implement a data strategy that delivers for customers, register to attend now.

Topics: Thought Leadership

Written by Tori Williams

Tori is the Content and Research Manager at Nimbus Ninety. She works with our partners to create content, research and events, designed to help our members grow their knowledge and spark opportunities.

Leave a Comment