Laying the groundwork for a modern data platform

Whenever a Datashift team works on a client project, one of our concerns is to deliver a solution that stands the test of time: a modern data platform that our client can rely on to answer all data needs for the following number of years.

Our strategy to achieve that is a Cloud-first strategy, which allows for a modular approach tailored to every individual client. Whether a client opts for an Azure or AWS implementation, we can easily integrate specific services and applications into the data platform and capture as many synergies as possible. Another benefit of our platform approach is that aspects such as data governance or data security are embedded (but we will still tweak it to how it works best for our client).

And while there’s obviously a lot of technology involved, we don't see Cloud-first as a technology story. Instead, it's all about how a modern data platform enables our clients to impact their business. Take the example of one of them, a digital media company. Their Marketing Department came to us with a demand to run more action-driven and segmented email campaigns while remaining fully in control (which proved to be the crucial part of their question).

Triggered by a concrete business need

There were 2 important aspects to our client’s demand. First of all, the efficiency of the Marketing Department's day-to-day operational work had to be maximized. People on the marketing team needed to be able to send out hyper-targeted email marketing campaigns without involving a data specialist to properly segment leads.

That is harder than it seems because anyone who sets up a business could be considered a potential lead for our client. Running targeted campaigns to a subset of those hundreds of thousands of leads, therefore, requires a modern data platform that provides easy access to various segmentation parameters: business area, geographical region, how long a company has been in business, or how a lead interacted with previous campaigns, just to give a few examples.

In addition (and here we touch upon the second important aspect of our client’s demand), the data platform needed to be well integrated with our client’s email campaign application.

Addressing existing bottlenecks

Through a series of in-depth discussions, we analyzed how those specifications could be translated into an effective solution and what kind of data architecture was needed to realize that solution. From the start, it was apparent that a list of email addresses along with segmentation categories wasn’t going to do the job. Instead, a more elaborate, hierarchical data model was needed to aggregate all information related to a specific lead into a single golden record and properly track conversion.

At the same time, it is important to note that an answer to the Marketing Department’s question was already partly in place. Apart from the limited capabilities of their existing data model, the big problem was that this solution involved a manual process with a lot of specialized knowledge in the heads of a limited number of people. Such a labor-intensive process left little room for experimentation or assessing whether email campaigns effectively targeted the right leads.

Keeping an eye on global, longer-term needs

Compared with their existing solution, additional components were needed to address our client’s longer-term needs, such as components that can handle a large variety of data in both batch and streaming mode, allow for scalability, and provide an application interface to integrate with external applications.

That became even more apparent as we noticed that many data-related initiatives were ongoing within the rest of our client’s organization. For example, an old data warehouse needed to be migrated into the cloud. Also, there was an initiative underway to realize a 360-degrees customer view, which had never really materialized. As a result, only 1 possible conclusion emerged. A Cloud-first strategy was the only viable way to build a modern data platform that enables our client to achieve their longer-term objectives for their Marketing Department and their entire organization.

Building a year-by-year roadmap

Our client needed a central landing zone, a modern data platform from which they can quickly launch and deploy applications to support their longer-term business goals. Obviously, you can’t build a modern data platform overnight. If you want to reap its long-term business benefits, you need to properly lay the groundwork and take the required time.

That’s why we worked closely with our client’s various data teams to build a global roadmap for the entire organization and define, year by year, the various components on which to work. Since a modern data platform is based on a modular approach, not everything is required from the start. Instead, this roadmap enabled us to define what to focus on first, in sync with our client's priorities and framed within a long-term vision and partnership with our client.

A modern data platform that puts end-users in charge

Moving to a modern data platform proved to be the deciding factor in putting our client’s end-users at the helm. Their Marketing Department, for example, can easily create lead segments for their email campaigns without relying on technical specialists and automatically pass along the segmentation parameters to the underlying data platform. From that point onwards, the platform will push the corresponding leads to the email campaign application, which will then process and feedback campaign results to the data model we designed.

Needless to say, eliminating a multitude of manual tasks also brings enormous time savings to end-users. As a result, they can spend their time on value-adding tasks such as testing out new hyper-targeted email marketing campaigns or refining campaigns depending on their leads’ interactions with previous campaigns.

Are you inspired by this story, and would you like to learn more about how to lay the groundwork for your modern data platform? Reach out to us. We’ll be happy to share more insights!