By Tuncay Tekle, PhD
The conversations around data collection, management and application seem inexhaustible, especially with the emergence of artificial intelligence and machine learning permeating nearly every aspect of the industry. What once centered upon data collection best practices soon shifted to conversations around privacy and data usage. Managing and processing data is no longer siloed within the IT department, ownership lies across the retail organization from corporate leadership to brick-and-mortar employees and even call center representatives. It seems the challenges with data are scaling as quickly as the data itself. And while the complexities are broad, there are four key themes that are dominating the conversations within the industry today:
- Multi-channel and multi-dimensional inputs
- New systems availability and compatibility
- Organizational structure & change management
- Customers as intellectual property
Multi-channel and Multi-dimensional Inputs
Opportunities for the acquisition of data seem endless. What started with mail-in forms, transactional history and email gathering has become a complex ecosystem of physical and virtual actions and behaviors across every engagement and sales channel. Mobile, geo-fencing triggers, direct mail/catalog, search and display ads, ecommerce/clicks/UX, point-of-sale, brick-and-mortar (traffic counters, RFID tags), streaming and video usage, virtual reality/meta-verse/gaming and engagement, marketplace platforms, third-party access, internal data and fulfillment feeds (ship-from-store, store-to-store, BOPIS, BOPAC, warehouse and inventory information), chat, call center, email customer services, social platforms, user generated content, end-to-end PLM, portfolio brand collaboration, payment services and more are all part of the new data collection universe. There is now a never-ending source of collection moments. Indeed, the challenge is no longer identifying data points, but rather, once gathered, how to feed these data points into a centralized system in a standardized way that allows for universal access and application within an organization.
Most companies don’t have systems with sources for endless data feeds. Keeping up with the changing business landscape for both data input and export that allows for seamless integration is among one of the top challenges for retailers today.
New Systems Availability and Compatibility
The shear increase in the amount of data and data feeds have fostered in new advances in how and where data is processed and stored. Retailers can achieve success now with both cloud and on-premise solutions. Cloud providers like: AWS Redshift, Azure Cosmos DB and Google BigQuery can be made both cost-effective and efficient with the correct implementation. By contrast, on-premise solutions like ClickHouse are also viable for processing large amounts of data at high speeds for enterprises with existing DevOps practices. The latter option can providefor better control and fast analytics but requires the engagement of an active development team.
The evolving landscape of the data world and the continued emergence of new technology and providers are driving retailers to evaluate their current systems across the organization. In many cases, however, existing systems are fully entrenched in the standard operating procedures of the business. Whether there are still components of Excel spreadsheets, home-grown legacy systems, server or cloud-based solutions in place, organizations are comfortable with their platforms and have inherent attachments to them. Entire teams have been trained on how to operate within their existing operational eco-system and know the glitches, exceptions, and patches in place. These systems are also extended into the business intelligence tools being used by disparate parties across the company and sources for multi-variant data are accessible sometimes only on one person’s hard drive (like weather history, retail calendars, and store profiles). Sounds antiquated? It is more common than not. Whether you are a digitally native brand expanding into the physical space or a traditional brick-and-mortar retailer expanding into digital channels, every retailer has its own set of systems challenges that have emerged only as a result of their growth and expansion. Layer on the speed of change within the data universe and the IOT (internet of things) seeping into consumer behavior (think fitness bands, theme park scan payment bracelets, implanted tracking chips, etc.) and evolution seems inevitable.
The need is there, but is the desire?
Organizational Structure and Change Management
Just because an organization identifies the need to address their business systems, doesn’t mean that change is easy (see new systems above). Shifting from legacy or static systems to new cloud-based agile solutions requires a full assessment of not only the data sources, but the data storage and data usage by all parties across an organization. Warehouse management systems, store systems/POS and inventory management systems are no longer exclusive to the teams that originated them. Similarly, planning, forecasting, merchandising systems and CRM and campaign management tools are not solely the responsibility of the product and marketing teams.
The historical nature of these systems being siloed by business area is what is now creating complexity and driving the need for an overhaul in organizational structure, talent roles and process procedures. Selection of a system is only the first part of the initiative; implementation and training are the critical steps for ensuring application, adherence and success. So, what is the path forward? Creating a cultural/readiness assessment is one critical first step to ensuring organizational openness. Consultants and/or project leads should create a roadmap starting with an audit of not only what types of systems are being used in their organization, but by whom they are being used and what the cross functional intercepts look like. Consideration also needs to be made for what, where, how and when data was and is being acquired.
Over time, organizations have compromised and kept old data in its original form in one system or area within the company and imported new data into a new platform separate from the legacy records. Blending data fields, flow and access, however, is another layer to the equation in modernizing workflow. Plus, considering how earlier data was acquired and whether or not it adheres to current privacy regulations, requires an assessment of if the data is usable or not. Legacy preference centers, credit card data and even early POS collected consumer data may or may not have been acquired with opt-in consent (implied or otherwise).
Customers as Intellectual Property
Ok, privacy. It seems you can’t have a conversation about data, collection, access, management, storage, usage without the word “privacy” surfacing immediately. The real solution to addressing privacy is readiness. Companies need systems and solutions that not only allow for multi-dimensional/variant data inputs and cross-functional, organizational-wide outputs, but that are capable of being agile and ready for changing regulations. Globally, we are seeing more and more regulations and requirements with regard to data collection and application. Consumers are the commodity and the manager of their own access. While younger generations are more transparent and forgiving with their data, maturing customers are more hesitant. Dealing with varying levels or degrees of acceptance adds some complexity to data regulatory enforcement. The laws, however, are the laws regardless of allowance. But laws vary by country, with Europe and states like California in the US driving stronger protections. Companies have to be ready to manage these geographic regulations and the inevitability that they will change. What used to be simple opt-in or opt-out choices has now elevated to product RFID tracking for carbon footprint adherence, duration of data use parameters, touchless tracking of consumer behaviors, and even less obvious data collection intervention like satellite data, vehicle data, application (phone apps and streaming apps) use and sharing.
But what good is data if you can’t use it? Not so fast. Personal data and individually attributed data are protected, aggregated data is more forgiving. With the use of machine learning and artificial intelligence, consumer clustering based on actions vs targets is more accessible to retailers. This pivot from one-on-one personalization to multi-level groupings and behavior-trigger-based actions is critical to optimizing merchandising and marketing efforts.
The Payoff
The continued complexities of data seem nearly overwhelming and perhaps insurmountable at times. That is why working alongside of practicing experts can guide you down the right path and make the audit, selection, implementation, and training process manageable. Who is doing it well? Overall, the industry is still catching up. The speed of change is outpacing retailers’ abilities to keep current. When companies commit to an evolving data practice, the payoffs are great. Brands like Decathlon in France who have applied RFID on all of their product are successfully eliminating friction across retail touches and throughout checkout with no scanning required, just basket-assessed payment.
Similarly, brands like NIKE are putting energy into data forecasting using advanced algorithms for accuracy that help ensure micro-level allocations, especially for new product launches.
Portfolio retailers that own a spectrum of brands are also gaining advantage as they can now collaborate their consumer and SKU data across properties, cross-selling and building greater consumer value.
Home brands like Home Depot are using data and technology to merge virtual e-commerce shopping with in-store physical shopping, showing e-commerce consumers exactly what aisle/shelf/bay a product is located in their preferred store for fulfillment options.
Amazon GO services are using behavioral data via product engagement and video observation to create more efficient and frictionless retail stores both from an inventory and merchandising preference and a no-touch check out experience.
The applications of data are endless with the true benefits being speed and accuracy in not only decision making but in consumer experiences as well. So, even with all of the complexities and challenges that continue to emerge in a data-driven business world, the pay-offs are much greater for both retailers and consumers who are that much closer to having the right products at the right prices in the right locations at the right time. Simple.
* Republished with permission from RIS (Retail Info Systems).
