Every company needs to be thinking about generative AI—here’s why
Prepare to see massive changes in how customers find products, engage with companies, and experience brands—thanks to generative AI. In some cases, these changes are already taking place.
Last updated November 3, 2023
Even if you’re not familiar with generative AI or large language models (LLMs), you’ve probably heard of ChatGPT, the remarkably human chatbot that can generate surprisingly conversational answers, passable college essays—even dad jokes.
While generative AI and LLMs aren’t exactly new, the excitement around ChatGPT has catapulted them to the forefront of global consciousness, launched an R&D “space race,” and sparked frenzied discussion about their potential impacts on information discovery, consumption, and content creation.
More than half of customers associate generative AI use with more premium brands, according to Zendesk research.
There’s no doubt that the impacts of this technology—particularly on customer experience (CX)—will be widely felt. In just a short period, we will likely see massive changes in how customers find products, engage with companies, and experience brands. In some cases, these changes are already taking place.
According to our research, nearly 70 percent of customers believe that most companies will soon be using generative AI to improve their experiences, with more than half tying its use to more premium brands. What this means for businesses is that thinking about incorporating generative AI into your customer journey isn’t a maybe, it’s a must—no matter how big or small you may be.
What are generative AI and large language models (LLMs)?
- Generative AI is an umbrella term that refers to an AI model that returns or generates an output based on an input or prompt. These outputs include text, code, audio, images, or video.
Many of the most popular and promising applications of generative AI are powered by LLMs. ChatGPT, probably the most well-known generative AI example today, is a chatbot-style prompt that allows users to query OpenAI’s GPT-3.5 LLM.
- LLMs are trained to understand language by ingesting massive amounts of text-based data (Open AI’s GPT-3.5 reportedly trained on 300 billion words from the internet). LLMs are incredibly good at generating, summarizing, or rewriting text, but they’re far from perfect. While text is impressively conversational, it can be factually wrong or based on outdated or unrelated content.
At Zendesk, we believe that AI will drive each and every customer touchpoint in the next five years. Even so, this is just the first chapter of a very long AI story. While it’s exciting to dream of where we’re headed, we must stay rooted in the knowledge that LLMs today still have some limitations that may actually detract from a customer’s experience. To avoid this, companies must understand where generative AI is ready to shine and where it isn’t—yet.
Here’s where customers expect generative AI to vastly improve their experiences
Customers believe that generative AI will transform the ways in which they buy from, engage with, and troubleshoot their problems with companies. Our research indicates that more than 75 percent say they expect it will enhance these interactions, no matter who they’re buying from. And enthusiasm only grows when they’ve actually had a chance to test drive the technology.
Among customers who’ve used generative AI:
- 78 percent predict it will soon play a crucial role in customer service
- Nearly seven in 10 say they will be more likely to buy from companies that use it in the future
As one tool in a larger AI toolkit, generative AI has the potential to level the CX playing field by making it possible for all companies to provide and scale higher quality experiences—all without needing to scale their budget.
Here is what most excites customers about generative AI (and what this means for your strategy today):
- High-quality, personalized engagement
- New ways to discover products or find tailored information
- Human agents with superhuman abilities
1. High-quality, personalized engagement
Forget having to fumble around for your order number or navigate a generic company home page. Customers want personalized service at every touch point, whether it’s in the discovery phase, the buying process, or any troubleshooting along the way.
Here are some of the ways that customers envision generative AI will improve their interactions with businesses:
- 77 percent of those who have interacted with generative AI want it to instantly generate unique how-to videos based on their specific questions
- 76 percent want it to instantly create unique promotions or sales specifically for them
- 61 percent want generative AI to communicate with them in a hyper-personalized way based on their past interactions with the company
Customers don’t have to be high value to get what feels like a higher-touch service experience. Though generative AI and LLMs will help in this effort, focus right now should be on ensuring that they can be deployed in a way that’s strategic and makes sense for your business—areas that we’re currently exploring here at Zendesk. Let’s be honest, no one wants a solution that may undermine customer experiences by not having the right security, privacy, and governance controls in place.
2. New ways to discover products or find tailored information
Generative AI has the potential to upend internet searches by delivering answers instead of website results. What’s more, customers can actually drill deeper by asking follow-up questions. Compare this with a traditional search engine where every additional query would require starting a whole new search from scratch.
Take this example: say you’re looking for a vehicle that will fit your family of four. By typing the question into a generative AI interface like ChatGPT, you could get a detailed description of types of cars (SUVs, sedans, etc.) that fill the bill as well as specific models. Budget is a concern, so you can further narrow the list by setting a price limit of $35,000. After getting a list of seven SUV models and their starting prices, you realize that you need one with large trunk space and a backseat that can fit a specific type of car seat. After a few more refinements, you quickly realize that the Honda CR-V is your best option for space.
In just under a minute, you have surveyed a wide range of four-person vehicles on the market, narrowed it down based on your specific needs, and zeroed in on a good option. Perform that same search in a traditional search engine and you get a completely different list, as well as a whole host of articles like “10 Best Family Cars of 2023”—not bad information, but not as personalized to what you actually need.
It’s a thrilling prospect—among customers who have used generative AI, 82% agree that it will become a central tool for discovering and exploring information in the future.
But it’s equally important to remember where we stand today. Generative AI can only pull from what it knows, which means that an example like the one above wouldn’t include recommendations of car models released in the past year. Besides the practical challenges of retraining LLMs, there are also legal challenges around privacy that must be overcome before this becomes a viable search alternative.
Product discovery isn’t the only area poised to change. Generative AI will also help companies reimagine how customers engage with help center content. Picture your chatbot receiving a question about how to process a refund, retrieving relevant answers from your help center, and then customizing a conversational response. Now pair that chatbot with Zendesk and add in the ability to actually issue that refund. That’s a game changer for your customers and your agents.
3. Human agents with superhuman abilities
Generative AI certainly isn’t the end for human agents. Instead, customers see it as a powerful tool for improving their interactions. Together, EQ and IQ join forces to ensure that customers reach the right person, issues are escalated when needed, and agents can provide better service with the right information (quickly) in hand.
It’s here— the elimination of manual workloads—where companies will realistically see the biggest gains from generative AI in the short term. Imagine an agent receiving an accurate, customized summary of a customer’s previous issues instead of having to dig up that information on multiple pages or systems. This alone would enable them to solve customer issues much more quickly and improve the overall experience.
3 in 4 customers who have interacted with generative AI want and are comfortable with human agents using it to help answer their questions.
While admittedly less buzzy than placing a grocery order or planning your next date night with a machine, customers agree. And they’re not squeamish about agents leaning on generative AI to make their lives easier. More than eight in 10 want generative AI to automatically send them to an expert human agent if it can’t provide the answer itself.
Part of that equation is the routing itself—understanding where to send them and who is available—but it’s also ensuring that the receiving agent has a summary of the information needed to get quickly up to speed on the specific issue.
Focus on immediate impact and plan for the future
Generative AI may be dominating headlines and attention now, but it’s important to plan for the road ahead. Sure, a chatbot that’s indistinguishable from a human opens up a world of possibilities, but AI is only as smart as the use case it’s trained for. If ChatGPT were a new employee, you wouldn’t immediately put them in front of a customer on the first day—even if they are great at speaking English. They wouldn’t know the ins and outs of your customer support operation.
Now picture AI that’s built on customer service interactions and, as a result, fully optimized for customer service. Suddenly that new employee understands the kinds of issues that customers commonly face and knows where to send them or when to escalate a ticket. And that’s where Zendesk’s focus is today—harnessing the power of these cutting-edge technologies in a way that makes sense for the CX use case.
A smart approach to AI will look at immediate opportunities for improvement—eliminating manual workloads and freeing agents up for higher value tasks—while identifying, testing, and fine-tuning additional intelligent layers to continually improve customer and agent experiences. These are the building blocks of an AI strategy that carefully considers where we’re at today with an eye for where we’re going in the future.