To compete in today’s retail environment, you have to be prepared to meet shoppers wherever they’re at. If the shopper is in store or on their way, the experience is fueled just as much by technology as the online or ecommerce experience is. Likewise, an in-store team member’s capacity to serve a customer “well” is more dependent than ever on quick, easy access to data and to powerful, reliable computing devices.

To win in retail, technology must be responsive and therefore responsively built.

Our journey to creating a ubiquitous tech experience at Target began with a simple vision to fundamentally change – speed up and simplify – how we access and deliver core technology for our engineers, our stores team and, ultimately, our guests.

For the engineers, that meant building the “easy path” – a one-stop shop to develop and deploy apps with high velocity – and enabling useful access to data and reusable components, no matter one’s level, device or location. For the stores and guests, achieving the desired experience meant freeing data from apps and enabling both data processing and compute at the edge.

Whether the goal is removing hurdles for an engineer or creating seamless shopping, inspiration and engagement for a guest, we knew we had to first internally chart a path to a ubiquitous experience.

What exactly is a ubiquitous tech experience?

Perhaps it’s best to first describe the opposite of ubiquitous – in terms of delivery and accessibility. And this is where we were a few years ago: We were batching data nightly from our mainframe, with multiple copies of data in a store; data was locked up in each application with no single place for developers to go. And, as a heavily outsourced IT organization, we were rigid in centralizing our data and restricting access. Our stores team had multiple, single-use devices and limited cloud-computing capacity. Our reluctance or incapacity to share information internally across our teams greatly inhibited our ability to deliver externally for our guests.

Today, we believe data and the capacity to produce, move or consume it should not be locked down. We expose nearly everything to an API. Likewise, our compute approach is to unlock fast, easy provisioning from the cloud to the endpoint device. Our app/data architecture is event-driven and built on microservices. Data and compute – and software deployment and management – are easily enabled at massive scale and down to stores’ multifunctional mobile devices, now in the hands of nearly every team member on the sales floor. Our engineers are empowered to make the right information (and the right amount of it) accessible to all persons (guests and team) and at the right time. Bottom line, they’re empowered to create value for Target.

How we built the ubiquitous experience

As mentioned in my talks at Google Next, the permission to reorganize, to change how we work, to innovate and take risks set the stage for today’s data and compute access and delivery. I can boil down our steps to ubiquitous and what I’d recommend doing by these three actions: removing technical debt, embracing open source and moving to the edge. Instead of listing all the details of these three broader actions myself, I want you to hear from some of our engineers who presented at Next on how we adapted the cloud (yes, Google Cloud Platform), to build for the edge, the data center, the cloud and do it all the same day. … Check out all the publicly available sessions.

Tom Kadlec is the senior vice president of infrastructure and operations at Target. Content in this article abridged from his two presentations at Google Cloud Next 2018:

  • “Our Journey to Create a Ubiquitous Experience”
  • “Democratizing Data & Compute”