By Mayank Baid, Regional Vice President, India, Cloudera
Fifteen minutes – that’s the limit of the average customer’s patience before he or she abandons the queue without making a purchase, according to a 2023 study conducted by Waitwhile. Globally, the average rate for e-commerce cart abandonment is closer to 75%, and in India the average cart abandonment rate stands at approximately 51%, as reported by India Marketers. Cart abandonment wastes the resources invested into getting the customer to that point of purchase. It is one challenge out of many that businesses face when it comes to retaining customers and reducing churn.
Common reasons for cart abandonment in India include concerns about online payment security, a general lack of trust in online retailers, high shipping costs, and lengthy checkout processes. Other challenges include difficulties in predicting customer behaviour and preferences, demand forecasting to optimise inventories, or delivering contextual advertisements that elicit the most response.
Customer retention is crucial because profitability can quickly erode with high turnover and customer acquisition costs. In fact, increasing customer retention rates by just 5% can lead to profit increases ranging from 25% to 95%, according to Frederick Reichheld of Bain & Company. In India, business priorities differ by sector and size, with large firms emphasising customer retention and competitiveness, mid-size firms focusing on growth, and small firms prioritising speed to market, while sector-specific goals vary across retail, construction, and metals.
Industry experts forecast that predictive analytics will account for 80% of AI applications in banking, particularly in areas like classification and categorisation. In the Indian banking sector, predictive analytics is revolutionising operations by strengthening risk management, providing deeper customer insights, and streamlining processes. Banks are leveraging predictive models to anticipate credit risks, detect potential fraud, and deliver personalised experiences based on customer behaviour. Technology providers play a key role in this transformation, offering solutions for real-time data processing and advanced analytics. For instance, some banks utilise hybrid platforms for fraud detection and personalised services, while others adopt them to improve customer engagement and streamline operations with tailored solutions. With bottom lines at stake, businesses must focus on enhancing customer loyalty by leveraging data to better understand and respond to customer needs.
Removing the silos, increasing access
In India, businesses focused on customer service have access to extensive data at their fingertips. Browsing patterns, sales, pricing, supplier orders, product information, and logistics can be leveraged to yield valuable insights on the customer. However, gaining access to these data sets is problematic as these usually reside across multiple locations or environments.
According to Cloudera’s 2024 State of the Enterprise AI and Modern Data Architecture survey, IT leaders (73%) recognise that some of their data exists in silos across their organisation and is not connected. Access to data is viewed as an insurmountable obstacle, with more than half reporting that they would rather get a root canal than try to access all their company’s data.
Eliminating these silos through a modern data lakehouse is crucial, without which, running large-scale queries or AI models to extract greater insights is impossible. Globally, several banks have recognised the need for a deeper understanding of customer profiles to cross-sell financial products and maximise engagement touchpoints. This enabled its data scientists and business users to leverage AI and ML tools to build new capabilities such as intelligent recommendations for products, resulting in a 138% increase in total bookings.
Making sense of unstructured data using AI
Unstructured data, the most abundant and fastest-growing type of data in India, can come in the form of feedback surveys, chatbot conversations, phone call transcripts, comments on social media channels, among others. High-quality predictions and insights mining call for discovery of new correlations, patterns, and insights from vast amounts of unstructured, semi-structured, textual, and relational data. This presents a challenge to data scientists tasked with managing and interpreting vast and complex datasets, as well as combining said datasets to derive a holistic, 360 view of the customer.
Some organisations encountered challenges with traditional data warehouses, which were costly and unable to meet the required speed or handle the scale of their data effectively. Today, technology providers are helping telecom companies develop a new analytical environment that could ingest massive volumes of real-time, granular network signal information and combine that with batch loads from billing systems, payments, and more. This is enabling teams to enhance customers’ mobile experiences and deliver relevant advertising.
Moreover, the advent of Gen AI has democratised access to analysis of unstructured data. Since Gen AI applications are easier to use and do not require technical skill sets to deploy, companies are now leapfrogging the traditional analytics cycle. They are bypassing the technical complexities of BI, data science and machine learning, going straight to deploying Gen AI applications that can quickly churn out deeper insights or practical actions that make business sense.
Speaking data’s language
Lastly, writing intricate SQL queries to develop machine learning models and crunch data sets can be overwhelming and time-consuming. Traditionally, generating SQL queries involves navigating the complexities of various databases, writing complex commands, and ensuring performance efficiency, and most businesses simply do not have this in-house capability.
This process can be daunting, even for experienced IT professionals and Accelerators for ML Projects (AMPs) help businesses address this challenge by simplifying AI adoption with a single click. Using AMPs, businesses will be able to deploy AI assistants to automate this task by translating natural language requests into precise commands, optimising queries, and providing understandable explanations of results. For example, the sales team can simply ask the AI assistant, “Why are sales down at this outlet? Will this trend continue? What actions should we take?” and take remedial actions to mitigate these issues. Teams across the organisation can also utilise AI assistants to establish clear metrics of success for its AI use cases. By tracking the right business metrics, businesses will be able to monitor its growth and trajectory alongside profitability.
The key to increasing customer stickiness and retention rates lies in actively listening to the data via AI tools and solutions. To get there, businesses will have to circumvent crucial challenges in gaining access to the data, preparing unstructured data for analysis and writing complex ML queries. Through deploying a hybrid data management platform, companies can explore building a comprehensive AI ecosystem, with all the data available at their fingertips.
