In the modern world of constant change, change is the only constant for every fast-paced business that must provide a compelling and engaging solution to its target market. There are limitations in the waterfall actualization model, such as the inability to effectively meet the timelines and rush that dynamic customer operations often present. But moving to Agile is more than just a method; it's a new style of thinking about the interactions between systems, processes and technologies in the organization.
In this paper, I present the views of how I participated in migrating a global beauty e-commerce brand from traditional e-commerce architecture to a more integrated approach governed by agile principles. But it wasn't just about adopting Agile -- it was about redesigning processes, integrating systems, and building up processes in order to create an environment that looks and feels seamless.
Traditional e-commerce architectures, CRM systems, and dedicated loyalty applications provide the main stack for e-commerce to deliver digital assets. Even if those systems worked, they were not adaptable to the constant loops of development cycles that are typically present now. It was clear, however, that deploying agile approaches in day-to-day practices did not address the more complex architectural issues arising from this embedded infrastructure.
Some of the critical pain points resulting from this were:
The solution required more than an Agile transformation; it required an engineering reconfiguration to offer the necessary tools for modular systems, CI/CD pipelines, and intra-team real-time collaboration.
The migration exercise for the client commenced by reviewing their current systems and ongoing projects. This included assessing the attributes of each digital asset and platform and their capabilities to support Agile integration while not interrupting existing project workflows. The solution was to introduce changes in technical improvements simultaneously with the deployment of Agile techniques so as not to disrupt operational processes. We made the following modifications:
Legacy systems were transformed and modularized to microservices supported by containerization platforms -- Docker and orchestration platforms. This made it possible to perform parallel updating and deploying which also accelerates delivery and minimizes disruptions at the system level while enhancing its scalability and fault tolerance.
APIs and middleware solutions such as MuleSoft and Apache Kafka were described as enabling connections between isolated, disjointed systems. These integrations allowed the sharing of data on the heartbeat and improved omni-station activity by plugging customers' e-commerce websites, clients' relationship management systems, or customer data platforms.
For CI/CD processes, tools such as Jenkins and GitLab were configured and utilized to deploy build and release cycles in accordance with the Agile approach. This enabled consistent and reliable deployment pipelines, allowing maximum manual errors and rapid development time frames because more time was saved in the processes.
Solving these basic technical problems allowed organizations to take full advantage of Agile while overcoming the problems associated with the traditional models.
In order to assist the client with such a significant change when many projects were already underway, we adopted the Scaled Agile Framework (SAFe) as it helps in integrating Agile practices within very large enterprises. We modified it to enable gradual implementation of Agile processes so that Agile teams targeted the most critical digital assets while continuing with other essential activities. This custom approach enabled us to roll out development more efficiently without interfering with other processes.
The SAFe is a comprehensive methodology designed to help organizations successfully implement Agile practices at an enterprise scale.
We began with mapping each digital asset, such as product display pages, recommendation engines, and customer profiles, to the customer journey. This mapping was crucial in identifying which assets had the greatest impact on user experience and business value. By understanding these touchpoints, we enabled the client to strategically prioritize assets for re-engineering, ensuring Agile integration efforts aligned with customer needs and organizational goals.
To create this robust digital product and platform taxonomy, we leveraged insights from established frameworks and research, following four streamlined steps:
This taxonomy enabled the creation of pods dedicated to specific functionalities and aligned them with overarching business goals:
Creation of Digital Assets and Platform Taxonomy revealed that ~30 pods comprising of ~270 full-time employees will be required to enable the front-end products for the e-commerce giant.
To ensure the organization was technically prepared for transformation, we focused on three key areas:
To drive the Agile transformation, we adopted and customized the Spotify Model to create specialized Agile pods, designed not just to be cross-functional but also aligned with the technical and business scope of the transformation, with a strong focus on mobile and web-focused digital products. Each pod archetype addressed domain-specific complexities while embedding Agile principles across the entire product lifecycle -- from design and development to testing and deployment of various digital assets and platforms for e-commerce. By customizing the Spotify Model, we established autonomous pods operating as squads while integrating a tribe structure to promote cross-pod collaboration.
This approach fostered alignment across customer-facing platforms, backend systems, and testing pipelines. It also allowed us to accelerate delivery timelines and drive consistent value creation, ensuring the Agile transformation aligned with the client's strategic goals.
In addition to their technical specialization, each pod archetype was tailored to the client's operational needs and designed to include a balanced mix of roles spanning product, digital, and technical domains. This composition ensured that every pod could address the unique challenges of its domain while contributing to the broader transformation objectives.
The Digital Assets Pod included experts in product design and development who worked collaboratively to deliver intuitive and scalable customer-facing platforms. Meanwhile, the Platform Pod supported the infrastructure needs of these systems, enabling seamless performance under dynamic business conditions.
To address resource gaps effectively, we conducted a detailed assessment of role distribution and expertise for the client, ensuring alignment with their strategic objectives. Drawing insights from industry best practices in resource planning and team structuring, we recommended retaining critical roles such as Solution Architects and Domain Subject Matter Experts (SMEs) in-house to safeguard long-term innovation and ensure continuity of institutional knowledge. For mid-tier technical roles like Software Developers, a selective insourcing approach was adopted, prioritizing project-specific needs to ensure expertise in key areas of development.
To optimize resource allocation further, we advised outsourcing standardized tasks such as Testing and QA engineering, leveraging external vendors to handle repetitive but essential work efficiently. This balanced approach drew upon frameworks outlined in resources like the PMI Talent Triangle and Agile team structuring principles, enabling the client to maintain agility while addressing resource constraints strategically.
To build agility within the organization, we focused on upskilling internal teams with targeted training on Agile principles, CI/CD tools, and API integrations. Certification programs were introduced to enhance technical capabilities, including container orchestration with the Certified Kubernetes Administrator (CKA) and cloud architecture skills to align with the demands of the transformation. Additionally, team members pursued Agile-focused certifications such as the PMI Agile Certified Practitioner (PMI-ACP) and Certified ScrumMaster (CSM) to deepen their understanding of Agile methodologies and enhance their ability to drive iterative, value-driven development.
This experience highlights the critical need to address both methodological and technical challenges in Agile transformations. For the global beauty e-commerce client, success depended on rethinking workflows and reengineering the foundational systems that support them. By implementing modular architectures, automating release processes, and aligning cross-functional teams, the transformation delivered faster product rollouts, improved customer experiences, and a scalable technology foundation for future growth.
Organizations embarking on similar journeys should treat Agile transformation as a holistic effort, balancing immediate wins with long-term investments in technical infrastructure. For insights into managing complex transitions, Large-Scale Scrum: More with LeSS by Craig Larman and Bas Vodde provides practical strategies for scaling Agile effectively while addressing organizational challenges.