The New Data Architecture


Andreesen Horowitz, a venture capital company, recently got people talking with a diagram it created that placed the cloud data warehouse (CDW) at the heart of what it called “digital data architecture.”

The model Andreesen published is increasingly regarded as the ideal way of storing and using data within an organization in technological circles. The model’s implications for data providers, as well as vendors vying to become clients’ preferred means of leveraging data for research, campaigns, product analytics, monitoring, and more, are perhaps the most intriguing.

This is particularly true for vendors whose solutions don’t fit into the ideal CDW system. Monolithic stacks, such as Adobe’s, that require proprietary storage, as well as replication-dependent analytics platforms, are examples of this.

The three tasks that client organizations engage in with data are: collecting, storing, and using it. The number of vendors that need their own storage platforms or data replication, a still-common necessity that new entrants are increasingly upending, has gradually complicated this relatively simple and linear definition.

The frequent lack of data standards and data integrity exacerbates the problem. Companies have profited from our passive compliance with data chaos to the tune of hundreds of millions of dollars in sales and billion-dollar valuations. These businesses have been placed on alert by the advent of the cloud data warehouse.

Clients are becoming more unable to work with yet another provider who has proprietary data specifications or replication criteria. It’s now simpler than ever to get started with a CDW, and there are a plethora of data governance resources available to assist with data quality issues. These methods should be used in conjunction with clear data requirements and enforcement across the organization.

Clients will ensure that each user is operating with the same, clean data by placing the cloud data warehouse at the core of the organization‘s data infrastructure and linking it to analytics software or other platforms.

In the future, vendors that can easily link to the CDW will have a much greater chance of succeeding than those who can’t — businesses that operationalize data but first need it to be obtained or imported and then stored in a proprietary database, for example.

We’re already seeing value change to the hundreds of companies that have sprung up in recent months to take advantage of the rise in CDWs — Hightouch, Census, and dbt, to name a few — as a result of the increase in CDWs. They also have the potential to produce results that, only a few years ago, would have required an entire team of engineers.

The architecture’s emphasis on the CDW opens up a whole new world of possibilities. Clients aren’t bound by a proprietary stack in which they can only use one vendor’s analytics or content platform

Although business value is shifting toward the CDW and the vendors that play well with it, clients are regaining ownership of both the data and the stack.

Follow and connect with us on FacebookLinkedIn & Twitter