Ganpath Thanumoorthy, Senior Vice President and Head of Customer Experience at Firstsource, explains what operations leaders need to know to grasp the opportunity behind the buzzword – and reap the benefits of early adoption.
Everybody is talking about AI, but if you look closely, you’ll see that not many businesses have actually deployed it at scale in their core operations yet. The reasons are varied and understandable: a reluctance to pioneer technology that’s considered yet-unproven, plain old inertia, or simply not knowing where to start. But I’m convinced: the golden window of opportunity for AI in operations is now, and nowhere is it bigger than in customer service. Especially business leaders who understand what AI means for their customers have the chance to leapfrog the competition. Here’s why.
The benefits of AI for customer service (beyond the buzzword)
Typically, without a business case, operations leaders won’t implement AI technology. Nobody buys AI simply for AI’s sake, no matter how big the hype. But the truth is that AI has been around a for a while, and has already proven that it can bring substantial benefits to customer service operations, among which are:
- Increased accuracy, as digital processing largely eliminates human error
- Improved productivity for agents, as repetitive tasks are taken off their hands
- Scale, as the routing of routine tasks to AI frees up agents to deal with complex issues – and enables businesses to do so much more with their existing resources, and reduces hand offs
These things are hugely important in a customer service environment, where response times and cost-to-serve are constantly being monitored, and where service expectations are always increasing.
But – while the cost savings and productivity benefits are significant – I believe that the real opportunity for businesses is in the unprecedented levels of delight they can offer to their customers today, driven by the recent introduction of large language models in AI. Early adopters have the power to set new standards for customer service.
How large language models are changing the AI game
Large language models (LLMs) are an enormous step change in AI development. With a corpus of billions of training data records underlying them, they’ve already absorbed the relevant patterns and key phrases, not just for standard requests, but also for niche and a plethora of domain-specific interactions. They can’t just produce a routine response; they can synthesise conversations and generate language that includes context and domain knowledge and is indistinguishable from that of a human agent. Plus, the responses draw from the combined knowledge that exists in the business, not just that of a single member of your team.
This is hugely significant because it means that your customers won’t just get faster responses. They’ll get better ones, every time.
And that’s good news for your business, too, in terms of:
- Agent empowerment: Even where AI doesn’t directly interact with the customer, it can shave off huge chunks off average handling time by providing your agents with automated call summaries and response emails, boost issue resolution with contextually relevant recommendations or drive revenue with personalised product suggestions
- Discretionary decision-making: the time, effort and cost associated with decision-making goes down immensely, as AI rapidly ingests, summarises and weights the relevant information
- Continuous quality improvement: With the help of LLMs, your QA isn’t limited to samples of 2-3% of interactions any more. You can QA 100% of all customer contacts, across all channels. This means your customer experience will get better all the time, at a fraction of the cost
- Better customer satisfaction at lower training cost: Agents don’t need the highest level of experience to deliver the highest level of service; the time it takes them to achieve required competency levels is significantly lower.
- The ability to create end-to-end automation: All of these capabilities – combined with data intake and workflow solutions – give businesses the power to create end-to-end processes that solve issues fast. They deliver speedy, truly personal and contextual customer experiences, and more varied and exciting tasks for agents who have been freed up to focus on individual cases.
Every service business is teeming with use cases
The use cases are plenty and go far beyond customer service. AI can pretty much enhance any process where time-to-response is crucial: just think issuing a customised quote for people looking to insure their property; think underwriting, where AI assesses dozens of documents within minutes, with a summary offered to an underwriter to make a final decision; consider HR, where job applications can be assessed and ranked by objective standards; social media, where a timely and empathetic response can make a brand’s reputation; or healthcare, where the ability to detect anomalies in an image with great accuracy can save lives.
I’m convinced: these things will eventually become table stakes across industries and departments – it’s only a matter of time. The businesses who seize the opportunity to improve service levels and outcomes for their customers now will have a competitive advantage over their competitors who don’t. It feels obvious to me that customer service – where solutions are mature, and the impact is immediate, visible and measurable – is the natural place to pilot AI in any business. And it’s a fantastic opportunity for ambitious operations leaders who are keen to move the needle for the people and the business they serve.
At Firstsource, we’ve helped dozens of organisations get started on their AI journey – from business case to technology selection, deployment and training. The outcomes speak for themselves – do take a look.
About the Author
Ganpath Thanumoorthy is Senior Vice President and Head of Customer Experience at Firstsource.