Executives from Neiman Marcus and AWS sat down at retail event NRF 2023 in New York City to talk about the luxury clothing retailer’s cloud journey and commit to a five-year deal with large-scale AWS services.
Two of the largest cloud providers attending NRF this weekend — AWS and Google Cloud — both noted an overall trend in organizations moving from a “lift and shift” transitional model to overall “transformation” as they engage more closely and at a larger scale with a cloud provider.
SEE: Hiring Kit: Cloud Engineer (TechRepublic Premium)
For Neiman Marcus, Vijay Karthik, CTO and senior vice president, was hired to tend this transition. He started in 2016 with just a couple of servers for mobile use. In the course of choosing a platform-as-a-service provider, he explored “lift, rearchitect, replatform and reboost” — starting in small chunks with programming done for the cloud from the ground up.
That required teaching programming teams to think cloud-first. Regardless of which cloud provider an organization chooses, they’ll have to consider how to rearrange tech teams, pace the change and keep up momentum during the transition.
“The teams need to be agile and the pace of innovation needs to be rapid,” said Sriram Vaidyanathan, vice president of omnichannel development and engineering at Neiman Marcus. “Pick the right tool for the right workload. We had to graduate to our level and be cloud native.”
Figuring out what the engineering teams would look like also involved getting input from e-commerce stakeholders, site merchants, and the Neiman Marcus digital and customer strategy team. As with any large transformation, executive buy-in was key.
Operational efficiencies come when you are cloud native, not cloud agnostic, Karthik said. The transition period from 2017 to 2021 — at which point Neiman Marcus entered into a five-year deal with AWS — taught them how to go farther.
Building a cloud team
Many people in retail will say customer needs are the top priority. When it comes to the cloud, Vaidyanathan said evolving and iterating on the way Neiman Marcus works with the cloud is part of building platforms and experiences that really meet customer needs.
“To actually put this in place you need to have the right foundation,” he said. “You need to have the right foundational team to be the custodians of the cloud.”
That will then propagate best practices across the enterprise before you start propagating the workspace across the cloud, he said.
Other recommendations: Keep your architecture simple, and don’t bite off more than you can chew. It’s important that people on both sides, customer and internal, don’t see the process or the result as too complex.
“Keeping it simple leads to building a repeating engineering practice,” Karthik said.
In terms of the in-house development team, ownership and empowerment is key to enabling people to move in a very agile manner, and excited, over-enthusiastic teams can complicate things. Instead, he continued, be mindful of how your engineering teams work, and keep them small and self-contained.
Other tips apply across leadership roles: Don’t micro-manage. Instead, set expectations and be a leader who can unlock their team’s abilities to fulfill those expectations.
Start small, but build from scratch
One of the important results of Neiman Marcus having the blessing of their CEO to experiment and change was that a lot of teams could pivot to focus on cloud. The Cloud Center of Excellence is a mixed, hand-picked DevSecOps developer team, but this team couldn’t be put in place until the organization knew they wanted to get started on their cloud journey, and they couldn’t really lean into that journey without the team with expertise. It took starting small but scaling quickly to succeed.
The platform engineering for the Cloud Center of Excellence team was done entirely from scratch, Karthik said. One of the good things about the cloud by nature was that they didn’t have to worry about patching — AWS did that. On the Neiman Marcus side, the next steps were to get Scrum teams involved and make good governance part of the team and company culture.
Platform engineering became a core competence for one sub-section of the Cloud Excellence team, Vaidyanathan said. They made sure everything was automated, worked in a new pipeline, could be scaled to be multi-region, and was available and consistent.
Lessons for cloud transitions
What can other companies take from this panel about how to run a smooth cloud transformation? Both Karthik and Vaidyanathan mentioned that automation was critically important. Amazon provided workloads for them. Anything that allows the developer team to code by hand less improves efficiency, and Vaidyanathan emphasized that code-level automation and automated quality checks make the process easier.
The Neiman Marcus developer team gradually became better at eliminating “process for the sake of process,” Vaidyanathan said. “If you have to move fast, if you have to actually have your team be able to have the rights to … ownership, you need to be sure there’s not too much red tape that prevents the teams from doing the exact thing they’re supposed to do.”
Failing fast is also important, he said. Company leaders need to understand and tolerate the fact that not everything will go according to plan every time and that the transition team will learn what to do better.
Along with having the tolerance for failure and failing fast, a team working on smoothing out a large-scale cloud transition should be sure not to make the same mistake over and over again. Instead, showing value to the wider organization is key.
Vaidyanathan also noted the importance of bringing in good partners if your organization lacks expertise in a particular area. For example, Neiman Marcus brought in a content delivery network partner for e-commerce.
Overall, the emphasis at the panel was to be deliberate about the speed of your cloud adoption — from a slow, exploratory beginning to the all-in deal Neiman Marcus and AWS have now.
In conclusion, the lessons Neiman Marcus learned over the course of their cloud journey included:
- Future-proofing: Find the right tool for the right job.
- Right-sizing cloud infrastructure: Changing from cloud-agnostic to cloud-native.
- Taking advantage of microservices, including microliths and versioning.
- Valuing availability over consistency.
- Emphasizing security and governance.
- Finding a good CDN partner.
- Utilizing knowledge from the experts.