Cloud Adoption in CG: Crossing the Rubicon and Thriving in the Land of No Return
Sri V. Raghavan
Sri Raghavan is data science and advanced analytics product marketing manager for Teradata.
For the longest time, cloud adoption in consumer goods companies has been the purview of information technology departments. Mostly. The decisions behind going (or not going) to the cloud often hinged upon issues like infrastructure optimization, flexibility in operations, performance and security and, sometimes, cost.
Currently, conversations about the cloud in the consumer product goods (CPG) industry are led more by the business. The focus is on enabling faster customer service, delivering product diversity, minimizing customer transactional costs and a general investment on maximizing end customer benefits. But the cloud, besides being an amorphous construct even today, is still proving to be a challenge when not properly understood and leveraged.
Important areas of cloud application for CPG companies include omnichannel journey, inventory management, customer service and supply chain optimization. The cloud is used to efficiently deploy and scale IT infrastructure and sales capabilities in new markets, channels and geographies.
The latest push for cloud is driven by the adoption of disruptive technologies like Artificial Intelligence (AI) to better serve customers. Real time integration of AI capabilities, either at the time of purchase or for in-store experiences, are best delivered through agile cloud platforms where the uptime required for scalable solutions is minimal. Cloud services are increasingly relied upon to deliver the integration of cutting edge in-store technologies with a unified digital platform that fully serves customers at all touchpoints.
Analytics in the cloud
Cloud analytics is the use of remote private or public computing resources to analyze multi-structured data at scale. These analytics involve the creation of a pipeline that includes data preparation, data discovery, feature engineering, model building, model evaluation and comparison, model operationalization and visualization. There are several advantages of doing analytics on cloud infrastructure, including:
Scalability and agility: As data sizes increase in volume, the need for a system that can vertically and horizontally scale, in equal measure without the loss of performance, becomes important. Scalability combined with an agile method of analytic inquiry, where each question builds sequentially on the previous question, is also important and the cloud becomes the perfect vehicle to facilitate that.
Unified data access: When data is accessed regardless of location, users are not forced to make do with partial data. The cloud is the perfect vehicle to connect to different sources and access all data.
Collaboration: Subject matter experts collaborate to provide their know-how on what data to collect, how the data is transformed, the kinds of analytics to implement, interpreting the results and communicating insights in an intuitive manner. A cloud environment reduces the time it takes to operationalize the insights, and increases the ease with which a wide group can become invested in the analytics journey.
Challenges to cloud adoption
While the benefits are many, the risks to cloud adoption appear in equal, and some might argue to a greater, measure. A 2020 Teradata survey (https://www.teradata.com/Cloud/Analytics) indicated that an overwhelming majority of respondents (83%) agreed that the cloud is the best place to run analytics. However, there are 4 distinct challenges to cloud adoption that CPG companies need to address:
Data Security: Data is central to an insight-driven world where every customer matters and every opportunity lost tends to have existential implications for businesses. When data security is compromised, the immediate ability to serve customers is destroyed and the long-term reputational effects are hard to recover from.
Lack of Expertise: Setting up a business on the cloud requires expertise that is consonant with cloud services and architecture. Cloud migration is often a dedicated stream of work that requires people with the bespoke knowledge and skills that are necessary to migrate workloads from on-premises environments to the cloud without any loss of functionality or continuity.
Compliance and governance: Regulatory frameworks are largely different across geographies. To operate effectively as a global concern the need to fit in neatly across the multitude of regulatory landscapes requires often a dedicated group of local professionals.
Vendor lock-in: It is no secret that today’s Cloud PaaS/IaaS providers are an oligopoly, and a powerful one at that. The top three vendors have the lion’s share of the market, so companies are concerned about how easy it would be to migrate workloads across providers should the situation arise.
Addressing the challenges and calls to action
While there are many challenges, companies can address them head on. Data security on the cloud is typically driven directly by APIs set up by developers, not IT operations. This makes security vulnerabilities more likely as all developers do not have security as a priority. Creating a separate focus where security teams are responsible for setting up access controls and data encryption makes cloud security a reality.
Resource issues in cloud management are quickly addressed by cross-training IT teams to adapt to the cloud. Retraining on-premises IT personnel with a cloud focus and hiring more cloud native practitioners makes it easier to embark on the cloud journey. Data governance and security are inter-related. Good data governance practices that can be instituted include assigning an SME for each data group, metadata tracking, developing policies for data integration and transformation and creating a compliance review of all shared data.
Finally, vendor lock-in is perhaps the most intractable of all the challenges given that there are very few cloud vendors available. Hybrid cloud environments where a part of the workload is situated on-premises and other workloads distributed across multiple clouds, mitigate these ill-effects. While this can increase the transaction costs of doing business, it manages unexpected contingencies with one vendor.
CPG companies are getting used to leveraging the vast resources of the cloud to deliver a strong data and analytics foundation that can address many use cases such as product bundling, fraud detection, pricing optimization, demand forecasting and supply chain management. Companies also need to be conscious of risks like security breaches. A planned approach for managing the cloud and creating a blueprint for migration will limit those risks and deliver on prospective benefits.
Sri Raghavan is Data Science and Advanced Analytics Product Marketing Manager for Teradata, with more than 25 years of experience developing products, marketing and sales initiatives that drive performance and profitability of organizations across industries and geographies.