メインコンテンツへスキップ

Q&A with Zendesk Chief Technology Officers

The company’s technology experts from around the world weigh in on driving innovation and shaping the future of CX

発行日 2024年7月2日

Customer experience (CX) has undergone a significant evolution in recent decades as a result of technological shifts — the telephone, internet, and Web 2.0. Now, AI is the next pivotal moment with potential to revolutionize CX more than any other preceding technology. Read on for Zendesk’s Chief Technology Officers perspectives on the latest developments in AI in their respective regions, plus preview what’s to come as automation extends across both business and personal bots.

Setting the Stage with Jason Maynard, Chief Technology Officer, AMER and APAC
Let’s talk about the business landscape – with the end of Zero Interest Rate Policy (ZIRP) as well as private and public markets putting more emphasis on free cash flow margins, it would seem possible businesses would feel pressure to forfeit on CX. Talk to us about how AI offers the opportunity to drive higher margins, but also remain focused on the customer experience.

Delivering great customer service was an easier proposition when capital was easier to come by. The primary drivers of great service, speed and quality have always been an important but expensive proposition, and despite a lot of progress on understanding the importance of service on customer loyalty, support functions are often viewed as a cost center.

Every CX team would prefer to have enough people to instantly resolve issues for customers, 24×7, and have compensation plans that indefinitely retain top talent, but unfortunately, that has never been possible and is harder than ever today. Fortunately, there are new opportunities to utilize technology to shift many of the constraints that have defined service operations for a long time. We’ve seen how LLMs are a combination of great communicators, and great problem solvers, the same skills you look for in great service team members. Although the best service agents still trump anything a model can deliver, there are many types of service issues where speed and cost are more important than a human touch. These are the areas businesses have an opportunity to drive high levels of automation into their service operations, freeing up human capital to focus on differentiated service where it’s needed, while at the same time delivering on the high pressure to improve margins.

Can you give an example that brings these concepts to life and validates where AI in CX can provide a return on investment (ROI)?
Of course. Here is an example for Universal Store, Australia’s hub for women’s and men’s fashion, that needed to centralize its customer support channels for seamless customer experience. With Zendesk, they were able to handle their growth with a smaller, but top performing team with automated processes like item returns, reducing their response and resolution times. As a result, Universal Store set industry best response times, enhanced customer satisfaction and retention while continually improving margins.

We see this across the board with customers who are adopting Zendesk AI, which in certain use cases can automate up to 80 percent of support requests and generates a 3x increase in immediate, automated resolutions. This contributes to a 30 percent decrease in resolution times and helps agents be at least 10 percent more productive. These numbers are why CIOs are investing in capabilities that have the opportunity to improve customer experience, while simultaneously delivering business and shareholder value.

What does this focus on AI mean for the role of humans? How should companies balance between AI and the human touch?
I start most conversations with customers by looking at the top categories of questions that drive the majority of their service volume. Those categories typically fall into 3 categories, issues where it is clear automation is the better customer experience, issues where there is a need for review because of judgment, and sensitive issues where you would always want a human. Oftentimes the highest drivers of volume aka. repetition are the areas that have the biggest opportunity for automation to be a better solution. This is where we start.

That said, many support teams also need to augment the skills and expertise they’ve historically needed to enable deep levels of automation. Support has historically been the lowest priority for technology support, behind sales and marketing, so CX teams often take these problems into their own hands. There will be an increasing need for more CX roles that are more technology managers, than people managers. Similar to the way the technical marketer role grew in the shift from channels to digital — CX will need team members that are part analyst, part system integrator, but still have a deep understanding of world class service delivery.

This will be a big transition for the industry, but it’s also an exciting time for CX practitioners to have a huge impact on how the business stays relevant in times of austerity. I get to work with insanely creative and customer focused people in CX organizations, that, with the right tools at their disposal, have a huge impact on their customers and business performance.

Staying Ahead and Moving Fast with Walter Hildebrandi, Chief Technology Officer, LATAM

There is a sense of urgency for CX leaders to show value from adopting AI. What metrics should they consider to effectively gauge the impact of such initiatives?
There are many, many metrics ranging from Net Promoter Score (NPS), Customer satisfaction scores (CSAT), Customer Effort Score, First response time, Average resolution time, First contact resolution, and more. However, it will depend on the type of business you operate.

For example, Magazine Luiza uses Zendesk to keep the focus on the customer and prevent bottlenecks and friction throughout the journey. To measure success, the company tracked the increase in both service team productivity and customer satisfaction. Another example is LATAM Airlines, which uses Zendesk for its internal employee operations, and tracks employee satisfaction as well as AI agent accuracy.

The industry is moving quickly. There is also urgency to keep up – if not, get ahead – of competitors. How is Zendesk staying across the newest improvements with LLM partners?
We are committed to providing the best generative AI experiences to our customers, and correct, the industry is moving quickly. Our existing partnerships with OpenAI, Amazon and Anthropic let us use the right generative AI to provide the best quality, speed and trust.

We recently announced a collaboration with Anthropic and AWS to help businesses deliver exceptional AI-powered customer and employee experiences. Then, in the wake of exciting developments announced at the OpenAI spring 2024 event, Zendesk is rolling out the new GPT-4o model to all of our Zendesk AI customers.

In fact, we were able to test and rollout GPT-4o for production Generative AI use cases within 24 hours due to our advanced Large Language Models (LLM) benchmarking and our flexible deployment model. This will improve performance and speed for bots, agents and admins at the same level of quality.

Lessons from CX Applied to Employee Service with Matthias Goehler, EMEA

Tell us about customer examples. Where might customers have high expectations when it comes to CX this summer?
While we all wish for smooth sailing in our travels, this is unfortunately not always the case. Travelers often run into flight delays, lost luggage and more, prompting the need to reach out to companies for support. When this happens, it is important for businesses to provide fast, AI-powered answers via self-service. For example, Skyscanner saw an 18 percentage point increase in customer satisfaction and a 76% drop in first response time after placing greater emphasis on self-service and rolling out a new help center.

However, some more complex problems need a human touch and AI can help identify those particular customers. For example, Eurail uses Zendesk to automate processes by using AI to read written customer requests for refunds that would normally not fit into the company policy, flagging certain terms like ‘emergency’ or ‘funeral’, and escalate them directly to a senior agent who has the authority to process refunds in exceptional circumstances.

Beyond CX, AI-augmented work is also making employee service better. Can you expand on this topic?
According to our latest research, a majority of Employee Service leaders believe AI is essential in boosting employees’ ability to tackle complex tasks, such as analyzing extensive consumer data to craft the best strategies. They think that AI has a bigger purpose than just fixing issues – it is an agent that turns challenges into opportunities for creativity, innovation, and organizational advancement.

Furthermore, close to 80 percent say that integrating AI into their workforce is not just an upgrade – it’s a strategy to carve out a competitive edge in the marketplace. By leveraging AI, teams enhance employee services with innovations in self-service, workflow automation, and training, leading to a more responsive and empowered workforce.

Closing Thoughts

Where is the future of AI in CX headed?
I recently spoke with Bloomberg about this future state of “bot on bot” where the whole customer service interaction will be automation on automation. We expect to see more automation within businesses and in our personal lives (check out Siri’s latest updates) in which a personal assistant can return a sweater – starting from finding the order number in your email to drafting an email to the retailer to start the return process, and eventually more. We are preparing for this scenario in which there will be LLMs talking to LLMs, and exploring what that infrastructure looks like across channels, products and APIs. – Jason Maynard

While AI is still in its beginnings one thing is for sure: The future of AI in CX will be transformative. Currently, AI is mainly used in real-time to solve customer problems. Looking into the future AI will help to predict potential issues before they even arise, enabling businesses to proactively address customer needs and ultimately drive their success. We will also see a bigger focus on data privacy. Considering we can expect more data privacy regulations such as the GDPR and seeing the needs of consumers, successful companies are recognizing that data protection and privacy can create a huge competitive advantage. CX leaders need to incorporate more privacy throughout the customer journey. This includes data masking, redaction, minimization, anonymization, visibility, field based permissions, among other tactics. Not only better privacy technologies are built but also AI can be used to support these efforts. – Matthias Goehler

AI is still in its infancy. Today, only 20% of support interactions are automated, meaning the vast majority, 80%, of requests still require humans to resolve them. But we predict in just five years, these percentages will completely flip with 80% of interactions becoming fully automated with AI, without an agent ever needing to be involved. These automated interactions will provide significantly easier, more satisfying experiences for customers with faster, more accurate, and more personalized support from AI. There will be less friction on every channel and integrated knowledge, data and systems to automatically take action. Looking even further ahead to 2030, we believe every single customer interaction will involve AI. – Walter Hildebrandi