While a lot of retailers have started their analytics journey and are at various stages of maturity, they are still struggling to identify the most appropriate model for sustainable analytics implementation. Retailers can adopt analytics across their enterprise using various options. This article tells you how…
PwC conducted a global survey to understand and compare consumer shopping behaviors. The survey revealed the following challenges for retailers, which can be solved using various data and analytics techniques.
How Can Fashion Retailers Adopt Analytics?
While a lot of retailers have started their analytics journey and are at various stages of maturity, they are still stru
ggling to identify the most appropriate model for sustainable analytics implementation. Retailers can adopt analytics across their enterprise using various options:
ORGANIC IN-HOUSE: The retailer augments its in-house analytics team gradually over a period of time and introduces analytics in a phase-wise approach. However, this is a slow process and the time to first insight increases drastically.
INORGANIC IN-HOUSE: The retailer buys an analytics company which can start working on various analytics projects across the enterprise. This approach can improve the time to value; however, there are integration issues and
the analytics resources may take some time to understand the organisation processes and culture.
PRIMARILY OUTSOURCED: Here, the entire analytics work is completely outsourced to an external vendor who ingests the various data feeds from the retailer, performs the analysis, and sends the results and insights back to the retailer. This approach can deliver quick results but in the long run can prove to be an expensive proposition for the organisation due to its as reliance on the external vendor along with loss of control on analytics operations.
HYBRID MODEL: This approach involves the development of in-house analytics capability through an external consultant who develops an analytics center
of excellence for the retailers and co-creates various analytics models with on their premises. This not only helps in gaining access to the latest best practices prevalent in the market but also keeps the control with the retailer at all times.
Analytics Framework for Fashion Retail
Fashion Retailers need to use analytics to generate in- depth insights across the value chain of their operations, including procurement, supply chain, sales and marketing, store operations and customer management.
Based on our experience in working with multiple retail organization, we have identified a retail analytics framework that can help them structure their programs in the four areas given below them structure their programs in the four areas given below:
- Supply chains
- Store operations
Marketing analytics can help to deepen customers’ insights, enhance retailer multi-channel performance, and improve the eﬀectiveness of their marketing initiatives.
MERCHANDISING ANALYTICS:Retailers need to use merchandising analytics to stock the right products at the right place and the right time. Merchandising analytics empowers planners to align their merchandising decisions with customers’ expectations. The key areas of merchandising analytics include assortment planning, demand forecasting and space allocation.
MARKETING ANALYTICS:To keep up with customers’ changing demands and ensure their loyalty, retailers now have the option of using marketing analytics to gain an in-depth insight in customers’ needs, have targeted interactions with them and provide improved services to them. Marketing analytics combine all relevant customer-related data from Point of Sale (POS) systems, CRM databases, loyalty cards, etc., with social media, weblog and channel data, enable customers to perform sophisticated analytics and share insights with them to help them optimise their marketing decisions. This can help to deepen customers’ insights, enhance their multi- channel performance, improve the effectiveness of their marketing initiatives and enhance their social media presence.
SUPPLY CHAIN ANALYTICS:The profitability of the retail is directly affected by the efficiency of the logistics function in the organization to maximize fulfillment of demand and avoid any back orders or stock outs. These include interventions in logistics, inventory and suppliers’ performance.
STORE OPERATION ANALYTICS: The performance of retail operations depends on various factors, including the effectiveness of the store staff, the cost incurred on them, reduction of pilferage in the store, management of inventory at the right level and improvement in the overall performance of employees in terms of footfalls and conversion rates.
Analytics Capability-Building in Organisations
Analytics is typically scattered throughout the operations of Retail organizations, leading to redundant costs and suboptimal adoption. They prevent organizations from achieving the scale necessary to completely nurture analytics talent and make the most out of their data. The first step is to conduct an internal assessment, based on the following:
BUSINESS APPLICATIONS: Understanding where analytics is currently being used and what is the maturity of analytics usage across an organisation’s functions
DATA: How various applications are captured, and internal and external data generated, and how
sophisticated is the organisation’s use of unstructured data
TECHNOLOGY: Leveraging existing technologies to consolidate data, storage, ETL, discovery, preprocessing, modelling and the quality of data management
PROCESSES: Verifying whether management and governance processes are optimal
Based on an organization’s performance in these areas, Retail organizations can ascertain their maturity in analytics and define roadmaps for their operations. The best way of realizing the benefits of analytics is by pooling a dedicated team of business analysts, statistical modelers and data analysts, and establishing an analytics centre of excellence within the organisation. It can be a one-stop shop for all analytics requirements across the various functions of retail organisations. This will help them establish analytics as a repeatable process and ensure that their best practices are continuously redefined and promoted. However, it is also important to understand that the relevance of the analytics department is dependent on adoption of analytics throughout the organisation and how quickly feedback is provided.
Value-Based Customer Segmentation In An Apparel Store
The management of the apparel store wanted to segment its customer base into five main categories, based on their overall customer value, to target customers accordingly. An FRM model was developed to analyse customer value where RFM stands for ‘Recency of purchase’, ‘Frequency of purchase’ and ‘Monetary value of spend’, respectively.
Once appropriate categories of each of the attributes had been defined, segmented were created from the intersection of values, which were then collapsed to emerge with five main categories of customer.
The resulting segments were organised from the most valuable (highest recency, frequency and value) to the least valuable (lowest recency, frequency and value). Customer segmentation helped the apparel store target specific group of customers effectively for various campaigns and promotions and allocate its marketing resources to the best effect.
Value-based segmentation helped the store identify groups of customers in terms of the revenue they had generated for it over a period of time, and thereby helped it calculate and allocate the cost of establishing and maintaining relationships with them.
Emerging Technologies in Fashion Retail
Emerging technologies need to be a core part of every company’s corporate strategy. C-suites are challenged to sort through the maze of technologies available to take clear-headed decisions about the most pertinent ones that will sustain their revenue growth and enhance their business operations. But with the torrent of technological breakthroughs affecting businesses of all types, it is important for executives to make sense of individual technologies and their application.
To help companies focus their efforts, we have analysed various emerging technologies to pinpoint essential ones we feel Retail organisations should consider. While needs vary from company to company, these technologies have been proven to make most significant impact across functions around the world.
IoT comprises a network of physical objects, including devices and vehicles that are embedded with sensors, software, network connectivity and computer capability, which enables them to collect, exchange and act on data, usually without human intervention. IoT is transforming everyday physical objects that surround us into an ecosystem of information that will enrich our lives. From refrigerators to automotives, IoT is bringing more and more products into the digital fold every day and is likely make it a multi-trillion dollar industry in the near future. This presents an excellent opportunity for Fashion retail companies to use sensor data to develop innovative products and provide unique services to customers.
Companies can now achieve their goal of ‘intelligence at the moment’, which will provide them with insights into and help them to analyse parts of their physical operations that were not measurable in the past. This data can be transformed into insights, delivered when and where it is needed to make and implement informed strategic and operational decisions, and in many cases, to gain a competitive advantage Some applications of IoT for retailers could be measurement of the effectiveness of their in-store promotions and footfalls across various categories. It could enable companies to gain relevant information on products that are in demand, identify ideal store layouts, predict stock outs in real time, etc.
The use of real-time POS data, especially when managed according to clearly articulated strategies, is reshaping how fashion Retail companies take decisions. A company may use AI to choose where it will expand its activity or to manage availability of products differently so that consumers are can have access to the products in high demand. Improved access to data by tracking inventory and demand-related planning can help to remove bottlenecks in the supply chain, direct investment in R&D, improve marketing and maximize the efficiency of the supply chains, which all work towards increasing profit the retailer.
Data-driven collaboration often includes sharing of insights on market trends and consumers’ buying behaviour. Professionals believe that such sharing leads to enhanced idea generation for products and trade promotions as well as more effective management of the workplace.
The most useful technologies for collecting such data are those that enable direct interaction with consumers. These include customer relationship management systems, Web 3.0 (which uses natural language searches, data mining and artificial intelligence technologies), online applications such as digital media campaigns and contests on social networking platforms.
Video analytics is another powerful tool for understanding the customer’s journey through a store. While video as a technology is not new, the maturation of computer vision algorithms has enabled automated tracking of objects appearing on a video. Moreover, with the application of parallel processing platforms such as Hadoop, these process-intensive tasks can be performed at scale. Combined, these tools enable retailers to understand the types of customers entering their stores, their precise movements and how they can direct the latter’s attention. Presented below are some capabilities retailers can utilise with video analytics:
- Automatic re-ordering: A change in the colour of perishable items such as vegetables can be detected by cameras and be used to re- order supplies.
- Traffic flow: Virtual tripwires can be used to calculate conversion rates from the sidewalk to the store and whether the majority of customers turn left or right on entering a store.
- Dwell time: The effectiveness of an advertisement or endcap display can be determined by tracking the percentage of customers who stop to notice an ad and the percentage that do not
- Demographics: Demographics help stores to gauge the age and gender of customers entering them
- Heat maps: These visually represent activity within a store and help to optimise its layout and sale of high-margin products.
- Queue analysis: Queue analysis determines the relative size of a queue and enables optimisation of staffing during peak and normal shopping season
- Security and safety: Machine learning algorithms can be used to automatically detect suspicious or out of the ordinary behavior in real time.
Challenges for Success
Irrespective of the nature of the solution being considered, there are some common deterrents that organisations must overcome to implement an effective analytics practice.
- QUALITY OF DATA: The greatest bottleneck in implementing a data-driven system is often data or the difficulty faced in accessing it and, its quality accuracy and completeness. The severity of this challenge is more pronounced in developing nations compared to developed ones. Many leading companies have begun using sophisticated data management systems such as Enterprise Data Warehousing (EDW) as a solution. EDWs not only facilitate streamlined access to data for analysts, but also integration of data across functions including marketing, finance and supply chain operations.
- DATA-RELATED BIAS: A data-related bias is another major impediment to accurate decision-making that analysts need to guard against. To illustrate this, let’s consider a marketer who plans to launch a health drink for the middle class working population in Mumbai. The marketer conducts a survey through a pre-launch survey, which covers respondents from households is an affluent locality on weekday afternoons, to understand the preferences of its target audience. Unfortunately, an analysis of such survey results would be inaccurate about the preferences of the target population because the majority of respondent are (1) financially affluent and belong to the upper/upper-middle class segments (location-induced bias) or (2) home-makers (3) or not professionals (time-induced bias). Therefore, it is critical to be cautious while analyzing such data, being aware of underlying assumptions in it to make sure of the usefulness of the results of the analysis.
- ORIENTATION OF MANAGEMENT: The other key criterion for successful usage of data in decision making is management members’ attitude to analytics in asking relevant questions about the data and basing their decision on answers derived from the ensuing analysis. In the absence of such an alignment, utilisation of insights is naturally below par, and funding of resources necessary for setting up a streamlined analytics practice assumes a low priority.
- OPERATIONAL & CULTURAL READINESS: Many organisations are still unable to adopt a new way of doing business, since their key processes are more attuned towards conventional methods rather than the use of analytics.
- AVAILABILITY OF DATA SCIENTISTS: There is a significant shortage of trained data scientists to drive analytical programs in organisations. Consequently, companies are facing significant challenge in finding experts who are skilled in data discovery, predictive modelling and statistical solutions, and have the required data- visualisation skills.
- READINESS TO SPEND: Use of analytical tools and solutions, along with the readiness to invest in building teams, requires continuous investments in time and resources.