Can there be a happy medium when it comes to AI and traditional customer support? The TechNational receives insight from global gifting marketpalce, Flowwow.
In the ever-evolving digital world, businesses are increasingly turning to artificial intelligence to enhance customer support services and drive growth. The e-commerce sector, in particular, is seeing a surge in adopting AI functionalities - 69% of customer support leaders are planning to invest further in AI in the year ahead. AI’s role in retail is projected to grow a staggering $20.05 billion by 2026, with giants like Amazon and Carrefour already using AI chatbots to improve customer experiences.
Despite the potential benefits of AI in customer service, not everyone is on board with this tech-driven approach. E-commerce companies are facing significant costs when deploying third-party AI software or developing their own. From a customer perspective, there can be frustration when receiving impersonal or inaccurate responses from AI chatbots. Nuanced enquiries often leave the bots confused and often lacking an appropriate response - not the desired effect when you might already be dealing with an unhappy customer. According to a recent study, 39% of UK shoppers are mistrustful of retailers’ and brands’ use of AI – increasing to 41% of Gen Z,
Flowwow, a global gifting marketplace for local brands, has faced the balancing act challenge when it comes to customer service. With up to 100,000 daily orders and 400 new sellers joining the platform each week, Flowwow’s customer support team handles around 1,500 enquiries daily, all managed within an in-built app chat. Eight support managers work in shifts, ensuring an impressive first-response time of less than 3 minutes and a 95% resolution rate. However, despite this positive outcome, Flowwow is always looking at smarter ways of working - striving to improve the customer experience while decreasing the number of enquiries.
Having heavily investigated implementing AI, Flowwow’s COO Vera Modenova shares three key insights for businesses considering adopting AI chatbot software:
1. A Good Website Should Cover the Most Basic Enquiries
Before deciding to implement AI into our customer support, we engaged with an AI-powered chatbot company. They conducted a diagnostic study of our support requests, revealing that the majority of Flowwow’s customer enquiries extend beyond basic FAQS, requiring more depth and insight than an AI bot can offer.
If a platform’s user interface and experience are designed using best-practices techniques and refined to your customer's needs, users will be less likely to contact customer support with common queries. In our case, the support team plays a dual role — they act both as customers and contributors to system improvements. Our support agents collaborate closely with the product team by identifying bugs and suggesting user-friendly changes thanks to their extensive customer knowledge.
A thoughtfully crafted UI/UX effectively handles the basic and routine user enquiries regarding the service, helping to diminish the overall volume of these types of queries. Tasks involving more intricate or delicate matters, such as managing complaints, resolving disputes, or handling emergencies, are delegated to human agents.
2. Cost vs Value Balance
While AI-powered chatbots offer efficiency when it comes to large-scale customer support, they might not be a viable ideal for all companies in terms of scale, cost, and resources.
For example - if you receive a large volume of user enquiries alongside 10,000 orders per day and have to maintain a staff of 50-100 operators, there are two possible approaches. The first one, as previously outlined involves adapting the product to address user questions. To do this it is essential to understand when and what kind of questions users are asking, allowing you to seamlessly incorporate answers into the product.
If the nature of your business makes this approach unfeasible, then embracing AI is undoubtedly worth considering. Take a tech giant like Amazon: if only 1% of users have queries, that still amounts to 16,000 enquiries out of 1.6 million daily deliveries. In this case, AI chatbots, coupled with automation tools and human supervision, contribute significantly to relieving the customer support team’s substantial workload.
AI-powered chatbots come at a cost: 23% of surveyed customer service pros believe that AI is too expensive to leverage. Some experts believe that the financial aspect of AI presents a major obstacle to Big Tech’s aspirations of AI being adopted across a spectrum of industries to make things smoother, save money, and improve productivity.
Successful AI implementation requires a dedicated product team. You need experts to train the bot, develop its logic and structure for providing answers, and enlist integration specialists. For AI to be as effective as possible, it must seamlessly integrate with all external services that the user interacts with. Additionally, when you introduce a new feature, the AI requires retraining.
If you turn to a third-party AI system, such as ChatGPT, the learning process will accelerate, allowing the bot to engage with your users not only about their experience with your service but also topics like weather, news, and trends. However, it is essential to estimate the cost of an external solution beforehand and remember that you will still need to train the bot specifically for your product.
3. Finding the Balance Between Human Touch and AI
44% of customer service pros who do not currently use AI in their roles believe artificial service may leave customers less satisfied than human communication. One concern is that AI might be unable to match the empathy and understanding human agents provide. Empathy, creativity, and problem-solving skills are critical when handling complex or sensitive customer problems. A robot can express sympathy and show empathy if you teach it to do so, but for a robot, it is currently difficult to independently distinguish in which cases this is truly necessary. Consider Netflix’s approach: their support agents are encouraged to infuse a personal, human touch into interactions with customers. This includes humour, personal anecdotes, and a warm, relatable approach.
However, a recent survey by Intercom highlights the evolving role of humans in customer support. Over three-quarters (78%) of support leaders say they expect AI to transform customer support careers in the next five years, creating new roles and opportunities. According to the survey by McKinsey & Сo, retail companies in the GCC countries have made the most progress, with 75% of respondents from that sector saying their companies have adopted AI in at least one business function. But, rather than diminishing human roles, AI and humans are forming a symbiotic partnership in the future of customer service.
The right implementation of AI and automation can relieve support teams of routine tasks and allow them to focus on empathetic customer interactions and creative problem-solving.
Expert Insight: Vera Modenova, COO at Flowwow has led the customer experience and operations divisions for 8 years alongside a team, developing a “glocal” (global + local) marketplace that brings joy to your loved ones around the world.
If you have a business insight you would like to share, get in touch with The TechNational team today or message us on X.
