Data Driven:
Migrating to AWS

This was a collaboration between Setfive and Data Driven. The end customer was a client of Data Driven's that we'll refer to as WF. WF is an affiliate marketing company which owns and operates about a dozen different website properties. WF works with advertisers to match users with offers and discounts as users browse WF sites. WF does this by combining behavioral, intent, and contextual data about their users and sites. WF was hosting their computing infrastructure in a traditional data center, which introduced several business and technical challenges.

  • Provisioning and installing new hardware took weeks and required a high up front capital commitment.
  • No managed infrastructure services were available, so even with the availability of open source options, WF’s small engineering team had to monitor and maintain a variety of services themselves.
  • Because of these factors, the engineering team was limited in the types of experiments they could run as well as the types of new features they could deliver.
  • Ultimately, these issues were becoming frustrating and unsustainable for the business team.

Initially WF consulted with Data Driven and Setfive to explore how to improve their developer experience at their existing data center. However, after an assessment of WF’s growth goals and their small engineering, team Data Driven and Setfive recommended that WF migrate their compute and data infrastructure to Amazon Web Services (AWS). Data Driven and Setfive worked with the WF team to migrate their entire infrastructure to AWS. Over a three week period we transitioned two production applications and two production databases with minimal downtime.

  • AWS’s "pay as you go" model allows WF to provision and deprovision resources as needed. Because of this the WF team is able to add and remove capacity during the day in response to traffic demands.
  • Managed AWS products like the Relational Database Service (RDS) and Simple Queuing Service (SQS) freed up WF engineers to focus their time on building revenue generating features instead of worrying about infrastructure.
  • Overall, transitioning to AWS has enabled the WF team to build product faster, grow their audience and generate more revenue.

WF’s move to the cloud was a boost to both their business and engineering teams. AWS was more expensive on a $/GHz basis, but the cost was offset by WF’s growth in revenue.

If you asked them if they would do it again you’d get a resounding "Hell yah!".