Diving deep into CX trends: a Q&A with Ray Wang

In our Zendesk Customer Experience Trends Report 2020, we shared our findings on the power of service in cultivating customer loyalty, unveiling the trends we believe will shape and define the landscape of customer service this year and beyond. To take a deeper dive into these critical concepts, we reached out to one of the sharpest minds out there: Ray Wang, principal analyst, founder, and chairman of Silicon Valley–based Constellation Research.

Loyalty is a major theme in this year’s CX Trends Report—it’s a driver of customer service, and it’s also a driver of success in your business. How does loyalty operate? To what degree is it a differentiator or a make-or-break?

Customers have been trading loyalty for convenience. They’ve been trading loyalty for value, they’ve been trading loyalty for status, they’ve been trading loyalty for anything. So the question, really, is how do we get them to be loyal? Can we earn their loyalty? And what does it take?

It seems to be the nonmonetary things that drive loyalty, whether it’s recognition at a simple level, access being linked to things or people, or products or services or experiences that other people can’t have.

The ultimate loyalty is being able to influence the product or service or offering that your customers are engaged in. This influence has to happen in the background and cannot be overt. In fact, there are different tiers of how loyalty manifests itself.

Moreover, that loyalty concept is being whittled, because customers are constantly being bombarded for things. It’s also because we can’t deliver on what we call ambient experiences—think Netflix, a great example of suggestions that are happening smartly in the background. It’s not in your face. It’s just, “Oh, would you like this?” And you just happened to be there. That moment of surprise, that moment of brand experience that pops up, that’s very special. And you have to preserve that—you don’t overburden someone with that.

That’s why this notion of loyalty is kind of weird. Companies often try to force fit certain types of models. For example, I checked in at a Fairfield Inn in Las Vegas. I didn’t want to see anybody. I wanted a mobile key. I just wanted to go straight to my room. That’s what I got—great. Not long after, I checked into a Westin Greenbriar, a nice, posh resort. Someone comes to the door: “Would you like a glass of wine?” Also great. I’m the same person; I haven’t changed. Sometimes I’m loyal to a brand when I consistently know what to expect, whether that’s getting a key and some privacy at the Fairfield Inn or the more personal touch of that glass of wine at the higher-end Westin.

The average customer does not care about terms like omnichannel and conversations—what they do want is what we’re promising these things deliver.

A customer wants to feel like they can communicate with you the way they want to, and they also want to feel like you’re proactively getting them, but not in a creepy way.

I think this is getting at this idea of ambient experience that you’re talking about.

Let’s put those terms, omnichannel and conversations, aside. What we’re really trying to deliver is a set of immersive experiences. Customers don’t care what department you’re in, customers don’t care what channel you’re in—as long as it’s the ones that they’re using.

And as we get into conversations, what we’re discovering, especially in support, is that what we want to be able to do on calls is, yes, experience a human on the other end. But as contact agents are analyzing the conversation on the call, what agents are rapidly doing are making suggestions. They’re providing next best actions, and recommendations. And they’re using that insight to help you resolve the case faster, or actually get the problem solving better. To actually relate with the customer better. Making sure that you don’t repeat everything all over again for the 40th time.

Those little things make a big difference. People will appreciate that. And if you’re really trying to get into chat bots, which is fine, just understand that by the time an issue reaches a human, it’s going to be much more complicated. Agents must be able to address complex issues.

The biggest challenge that we have in delivering this level of mass personalization at scale and ambient experiences: We have to determine when we automate something, when we augment it with a human, when we augment it with a machine, or when we actually automate it completely. What’s more, there’s a spectrum of these types of experiences, and people are going to have different preferences.

It’s critical for companies to do sensitive, on-point, and proactive communication with their customers. This involves collecting data and properly interpreting it—like you say, everybody’s different. Some people are going to want a text message, and others launch their phone across the room and say, “Why is my former favorite chocolate truffle company texting me at 2 in the morning?”

But as soon as we start talking about that kind of data, I think about the privacy scandals of the past couple of years. Are they behind us?

I don’t think this issue is in the past. Authentic brands require trust. Your experience is based on that trust level that you have. So, whether you hopped onto self-service, or whether you engage with a bot or with a human, if trust doesn’t exist, there’s no way. A good experience is impossible because my expectations are set in a perverse way. So how do we fix that? One aspect of the challenge is making sure your brand is consistent in its interactions and sets a consistent expectation.

Inevitably, at some point, a company breaks trust—a software outage, say, or a data breach. We saw in this research that many people will ditch a company after one bad experience, but after multiple, most people will switch. Is it possible to regain trust once it’s lost?

Yes, I think that’s possible. I mean, you’re not going to win every customer. And I don’t think we necessarily have to try to win every customer. You do want to win your high-value customers, and do treat them a little bit differently. If you’re losing high value customers, that’s messed up. That, we’ve got to go fix—the cost of customer acquisition is high.

But it happens.

Oh yeah, it happens a lot.

I think segmentation is important. They’re going to use tools like an AI to figure out which folks you want to keep, which folks are high value, which folks are high probability prospects, what other relationships are in their network—a lot of different factors. You want to understand the relationships but you don’t have complete data either. So the goal is not to have a customer angry at you, but the goal is to make sure that when they’re angry at you, that they’ve got a good resolution or there’s a good outcome at the end.

Customers expect all kinds of things: self-service, live chat, and messaging, and communities. And less than 30% of the companies that we surveyed in the Zendesk CX Trends Report 2020 report are offering these things. Do you have a take on why? And how steep is the cost of that?

Many times, when people aren’t leveraging those technologies and other types of interactions with their clients, they may not know how. From a technology perspective, they’re afraid. They’re afraid it’s another cost. They’ve been trying to drive down the cost of customer support and customer experiences, and what happens is that they’re realizing it’s actually everything’s causing things to cost more.

Part of it is a lack of automation. If there are things you can automate, it’s more than deflection. It’s automation. Which is important, but it isn’t going to solve the problem

Let’s say we get to a good level of deflection. And we automate a lot of the responses. The only time you’re actually going to talk to a human is exceptions. Instead of $10 an hour, those folks are going to make like $20 an hour. Or instead of $5 an hour, they’re going to make $15 an hour. Because you’ve got the higher level support agents.

On CX investments, to me it’s very obvious, but it’s not always well funded. And I think that’s the challenge. And a lot of customers need to know why we’re doing this, and they need to be able to justify why they’re doing this too. With a limited budget, they can only make so many choices. And that’s really the answer as to why there’s only 30% on some of those channels—some companies are just trying to perfect contact centers or an e-commerce channel. And that makes it hard.

How will this changing landscape of customer service affect customer support as a career option? Some of the agents I’ve spoken to say, “Look, this is what I want to do for life.” And others see it as just one step on an interesting road.

Oh, I think careers are going to change. You’re going to be training the bots. You’re going to be making sure that these knowledge bases are good. You’re going to be making sure that these AI systems aren’t discriminating or have unwanted bias. There’s going to be a lot of work! To get the precision decisions is not easy. And to rapidly get to precision decisions is superhard. Because we’re connecting on the fly that, “Oh, I didn’t know you own these three other things that are related to my companies.” Or “You have other interactions with me as a company.”

For instance, “I didn’t know you have DirecTV, and you’re an AT&T wireless customer, and you have something else.” Like a phone line service. Sometimes people don’t see all those things either. And so we’re going to have to piece together stuff on the fly. Also a factor: Sometimes customers don’t want you to know that.

What’s your take on how AI is transforming customer support and customer service generally?

Right now we’re in the learning stage. We are training these systems, and we’ve got a lot of work to do to get the basics right and improve precision decisions. The interesting thing about voice and voice navigation is when you’re using it, it’s not a tree-based navigation. It’s basically built around the customer’s problem. And so you have to design things differently. As a customer, I’m not going to wait for 13 different prompts. That’s not going to happen. So you’ve got to get it right the first time, which is really hard. People are going to be very disappointed when they ask a question and they don’t get the answer that they want. Quite simply, if you don’t get it right on the first time, you’ve lost my trust. So when you launch this thing, you had better be good straight out of the gate.

The second thing is, as we get some more complicated questions, the question is, “Well, how do you get this to work? How do you make sure that this is an easy way of connecting folks to the next answer and the next best answer? And am I able to learn?” So we have to build these systems to learn. If we’re not building to learn, we’re kind of wasting our time. Because if you lost an opportunity to understand the context of it, of an interaction.

And in the digital world, what’s amazing about the digital world is that we have feedback loops. Every action that we take, and every choice that we make is a digital feedback loop. That’s the trivial. We know what time of day, we know what process, we know where you were, we know what location. Sometimes we might even know what your heart rate was.” And when we take that context into action, we can then figure out, what choices were made, why choices were made. And get it down to that level of detail, which is great. So we can learn that. “Okay, this person, at 10 in the morning loves to talk to a person. At two in the morning, now they want to use chat bot.” So even just channel preferences we can figure out.

We are noticing a trend of more and more customer experience executives in the C-suite, such as the chief customer officer. Real trend, or flash in the pan?

Oh, the chief customer officer has been growing. We see a lot of the decision making, especially for things like SaaS, is happening between the chief customer officers who are focused on customer experience, and those who are on the IT side. There are two scenarios: “OK, I have to replace a contact center. Great.” And then: “I have to change my customer experience.” Which is very different—that’s coming through the business side of the house. This is why the chief customer officers of the world are so critical: They are the folks who are trying to change the world in terms of their customers’ experiences.

We all agree that everyone should be thinking about the customer, but who is backing it up? Who’s actually delivering on the programs? Who’s changing the way the engagement works? Improving the channel? Making sure the channels all work with each other? That’s got to be the chief customer officer, trying to figure out, “How does it work? What’s my journey look like? How do I orchestrate that?”

We’re also seeing that customers expect companies to collaborate internally. They don’t want to be bounced around from person to person.
Yeah, customers don’t care what department they reached.

Is that trickier than it sounds? If you’re asking departments who don’t interact much to do so seamlessly sometimes in the name of the customer?

It doesn’t have to be. When you’re not focused on being account centric, and you’re not in a situation where people aren’t trying to solve a problem but are turfing an issue to someone else, you have a situation where we have a crew first, customer second culture. When you’re not focused on being account centric, and you’re not in a situation where people aren’t trying to solve a customer problem, then yeah, it’s a different thing when people turn things over. But I think in most cases people try to work with each other. They just don’t have the information. Give them the information as the core first step, so people know what the customers are experiencing. Then they have an idea what their interactions have been like. This is all very, very important.

So if I don’t know the interaction history of that customer, or if I don’t know how to resolve a problem, or all that data has to be repeated over again… Yeah, that’s massive friction and keeping people from collaboration. But if we’re all looking at the record at the same time—respecting privacy laws, of course—people are willing to act. Having information access is really important in collaboration. And then having rules of engagement and how you handle that customer is the other piece that has to follow.

What do you feel is the optimal number of tools or outlets that companies should be providing for their customers to engage with them? Is there a magic number?

Oh God, you should have self-service. You should have a chat bot. I should still be able to talk to a real person and it should be 24/7. And I should be able to have a knowledge base I can hit, or a community that can answer my questions for me. But I’m on the extreme.

When one company takes a leadership role in terms of offering customer support and experiences that transcend the status quo, what impact does it have industry-wide? Does it create an arms race, if you will, for the other companies in that industry to provide something similar?

Oh, definitely. I mean, think about chat bots this year. How many Alexa demos did you see this year? It’s crazy. And so everybody’s like, “OK, we’ve got to go do this. I don’t want to miss out on the channel.” And it’s true. It’s the same thing when people were looking at social channels, or at self-service. One company goes where they can differentiate—and then after that, everyone expects it. You don’t have a choice… I mean, you’ve got to catch up.

The trick is to get the early-adopter customers to do it, do it really well, and then everybody will follow.

Check out the Zendesk Customer Experience Trends Report 2020 for must-have information on the current customer experience landscape.