The Emotional Labor Crisis in CX

In this episode of CX Now, host Lauren Gold, Chief Customer Officer at Kustomer, sits down with Jess Jackson, who Lauren describes as a major trailblazer in the AI space. Jess inspires with thought-provoking questions and a new way of thinking through some of the biggest challenges in CX.This episode’s topic: the emotional labor crisis in customer experience.

This interview has been lightly edited for clarity.

Lauren Gold: I'd love for you to share with the audience what we're going to talk about today and why this topic was so important to you.

Jess Jackson: Yeah, for sure. One of the things we had talked about was what I would call an emotional labor crisis in CX. The reason it's important to me is because I care about my people, but also I was an agent myself. I've come up through the ranks and I completely understand it from a first-person perspective, not just from an academic perspective.

I think I have a fairly unique perspective of understanding both sides of the coin — the leadership and business side, what drives that emotional labor crisis, and also what it actually feels like to be somebody who's a part of it. So let me define what that is, because emotional labor can mean a lot of things to a lot of people. And "crisis" is obviously a scary word too.

For the purposes of this conversation, emotional labor is when you need to regulate both your own emotions and also manage someone else's emotions — in this case, as part of your job, at scale, and specifically on demand. That's the biggest component of this, because CX is one of the very few places where emotional labor is continuous. You've got to do it for eight hours a day. It's not something that's occasional, that you kind of flip in and out of. It's scripted — you don't get to choose when or how you're going to empathize with people. And it's measured, which I think is one of the things that makes it more unique versus the emotional labor you put into your personal relationships. You're measured on things like QA, CSAT, handle time, and so on.

It's inherent to the role of being in CX and being a frontline agent in particular. But the reason there's a crisis aspect — for want of a better word — is that it has deep complexity to it. There's an emotional intensity to contacts that is increasing due to AI and self-service, as they strip out all of the easier interactions that people used to handle. There's also economic pressure pushing for higher volume per head and stricter metrics, which results in thinner staffing as well. The crisis exists because unfortunately most companies still treat emotional labor as a very infinite and somewhat free resource — when I think it's anything but.

Lauren Gold: Thank you so much for sharing that perspective. It's definitely an important topic, especially in this AI transformation we're all going through. I think about how our teams were structured in the past and how some organizations are changing that now. Let's talk about the previous, most standard way teams were structured — and why does it make the work so exhausting?

Jess Jackson: I want to say one thing before I get into this: I don't inherently view AI as good or bad. It's a tool, and just like any tool, it can be bad when used in the wrong ways and excellent when used in the right ways. But more and more, we have bots and self-service deflecting the more simple — I'll call them emotionally neutral — queries. What's then left for the actual people who handle the remainder are your edge cases, your failures, and the more emotionally loaded scenarios. Where it might not be complicated, but a customer thinks: I want to speak to a real person because I'm upset, I'm going through something, and I don't really want to hear from a bot right now. Or things that are escalations that your self-service channels simply can't fix.

The way this is really changing is what I'd call a complexity ratchet. The average interaction complexity and emotional intensity are going up, but metrics and staffing models aren't necessarily adjusting to that. So you've got leaders who are still quoting old benchmarks for handle time and occupancy that assumed a very different mix of work.

If you call me with something simple — "I lost my credit card at the pub last night, I've looked at my online transactions and I can see that no one's used it, I just need you to cancel it and reissue" — that's probably not complex or difficult, very in line with your processes and procedures, and probably not very emotional. For an agent, that's a downtime call. Simple, easy to breeze through, doesn't require a ton of emotional investment.

That conversation is also ripe for automation. So that conversation is now being removed from your agent. What the agent gets instead is something more like: "I've lost my credit card and I can see mass amounts of fraud on this card. I'm also about to default on my rent and I need your help desperately." That first call is not going to be anywhere near the same length as that second call, even though both of them are, at their core, "I need you to cancel my credit card and replace it."

When you have metrics that don't change — handle time, QA, CSAT — and CSAT is also much easier to earn on that first call than the second — and those metrics don't adjust as you move the simpler things out, it causes a real difficulty in the complexity mix.

That said, there are new opportunities too. AI can handle your rote explanations, policy lookups, note-taking, summarization — all the stuff that, frankly, agents don't want to do anyway because it's boring. So your humans can spend more brain and honestly more heart on the customer. AI can also surface early warning signals — spikes in conversations about fear, confusion, or anger around specific issues — and send that to your product managers or escalation teams.

But the risks come if you're using AI purely as a cost-cutting wedge. I want the same headcount, double the volume — or the same volume, half the headcount. That accelerates burnout, drives higher turnover, and you lose internal IP and your best agents. Things are shifting, but the metrics themselves haven't caught up with the change in contact mix.

Lauren Gold: Yeah, that's a great call. We really do need to reimagine the metrics of success and what we're holding agents accountable for — especially as more time, more empathy, and more complexity are going to increase average handle time, while still potentially producing great results for the customer and the brand. You spent so much of your career in the banking and fintech space, where you faced this every day and had to bring your agents along with you. How did you arm your agents to be prepared for those complex conversations, and how did you prevent burnout?

Jess Jackson: I think it starts with a mix of diagnostics and just staying close to your team. It really is as simple as listening to your people as humans and adjusting as you get feedback.

Some quantitative red flags to watch for: high turnover, of course — but I actually think high early-tenure attrition is probably the most important. If people are quitting within three to nine months of a role they've just been trained for, something is up. Similarly, escalating sick leave, mental health leave, or people reaching out about EAP and other resources to get through the day — those should be massive red flags.

One that people probably don't think about: if your CSAT is relatively okay but your internal engagement metrics — your ENPS or whatever you're using in the CX org — are bad, that mismatch means your agents are likely overcompensating for the emotional load. They're keeping your customers okay, but they are taking on way too much themselves.

The qualitative red flags are equally, if not more, important. If you have people saying things to you like "I'm numb now," or "I can't listen to my friends' problems after work," or going silent on tougher topics — sometimes agents get treated as therapists without the right training or backing for that. If you're hearing these kinds of things, you need to think about how to best equip your team to not take on everything they're hearing and get worn down by it.

The other thing I look out for is management doing off-the-books emotional triage — late-night DMs, one-on-ones that turn into counseling sessions. I don't know anyone who has felt like they had to let it all out to their manager and then just felt better, everything is great. Getting to that point means we've reached a critical inflection point.

In terms of leadership behaviors — the trap a lot of people fall into is celebrating heroes who push through the impossible, like a massive queue spike, without changing the actual system that required those heroics in the first place. That can be a real problem. Being celebrated for getting beaten down while continuing to get beaten down isn't as great a feeling as some leaders might think when they say "good job."

And the biggest one: when you're having discussions about burnout and emotional load, be real about them. There's nothing worse than responding to someone's pain with "I'll send you to this e-learning" or "please call EAP." Ask yourself why, and what's going on, and how can you actually help the person in the seat.

Lauren Gold: If a business finds itself in that situation, what can a team lead or a company do culturally to address it?

Jess Jackson: The first is the obvious one: redesigning the work and the metrics. Calibrating your metrics for complexity. Creating different handle time expectations and throughput expectations for high-emotion or high-risk conversations versus simpler ones is really important. Not all contacts are created equal, and that disparity is going to get greater and greater as automation continues to charge on.

Another way is around balanced scorecards. Don't just measure efficiency — efficiency is the easy one, keep it — but balance it with wellbeing and quality metrics. What was the case resolution quality? What was the repeat contact rate? And what was the peer rating of the emotional load on that call? Was this actually quite difficult for this agent to handle? A balanced scorecard will look different per org depending on what you're dealing with.

Then there's the ability to decompress and rotate. Build in micro-breaks designed for after heavy contacts. Have time blocks for people to actually come down and decompress. Once upon a time, back-to-back calls were a combination of easy calls — your downtime — and more complex calls that you steeled yourself for. Now, if they're all complex, doing that back-to-back-to-back for eight hours is a lot. Rotate people through lower-intensity work by design, if you can.

The other big thing is building actual emotional infrastructure — not just classic perks. Train your agents and managers in de-escalation and boundary setting. That's a big one. They don't have to take on everything that's coming at them. Recognizing customers' trauma responses and not getting sucked into them, and knowing when to escalate for safety. Don't penalize people for escalating contacts that really should be escalated. Teach them how to care and be empathetic without carrying everything home with them. That comes from things like clear escalation and safety protocols, having playbooks for uncomfortable situations that people don't want to talk about, and building channels to involve specialists where appropriate — so your agents aren't improvising on the really big things, because that's what causes the biggest stress.

The last thing I'd say is having leadership rituals that normalize reality. Regular load reviews. Sessions where you're looking at not just your queues and SLAs, but the emotional load — what type of stories are we hearing more of? Do we need to redesign our products? Do we need to redesign our processes? Do we need to create resources for our customers to get help?

Have story-based feedback loops from frontline to leadership. Call listening sessions — anonymized call snippets or case summaries that actually show your internal teams the emotional reality of the policies they produce and the product decisions they make.

And transparent communication. Honest discussions about limits. Where can the company actually flex? Where can't it? What's a rigid policy because it's regulated? And the why — most importantly. Teach your people business and commercial acumen so they can actually make decisions and they're not constantly negotiating against impossible expectations, because that's a huge load for them as well.

Lauren Gold: So many good insights there. I want to shift gears a little bit, because sometimes in the scope of this conversation we're talking about AI taking the easy conversations away and the load going to agents. But what about the more positive spin — where is AI actually supercharging agents and improving their experience, and therefore the customer experience too?

Jess Jackson: I think AI can do a ton in this space, actually. And it's probably one of the most criminally underutilized tools in this regard.

An obvious way is having AI look at what kinds of interactions your agents are having and provide high-level information to your executive teams and decision-makers: this is what's going on in your customer base, this is what needs to change — in products, services, and policies — before a customer even reaches out. Remove some of those difficult conversations altogether. That's good for the business, good for your customer, and good for your agent.

Another way: when I was an agent, it was rough. It was a long time ago and the technology wasn't where it is today. I can tell you that when a call drops into you and the first words out of the customer's mouth are, "I've already explained this issue four times to other people" — now you're dealing with the ghosts of conversations past. You weren't part of those conversations, you don't know why it was difficult, you don't know what happened, and the customer is already annoyed.

Eliminating that — by having a single conversation view that actually analyzes the history — makes a real difference. Not only has this customer called about the same issue five times, but it's increased in severity: they've gotten more agitated, they've used phrases like "you don't understand what I'm asking you to do," or even asked to speak to a real human, which people do sometimes say to real humans as well when they're too policy-focused. Being able to analyze those things and put them in front of your agent is actually easier for everyone. I can say to a customer: "It looks like you've called us a few times about this. Do you mind if I take a second to go through your history so you don't have to repeat yourself? I want to make sure I get this right for you." That's immediately better for the customer and better for the agent, and it gives the agent the context about what went wrong in the past.

The least exciting part of being an agent is after-call work — taking notes and updating admin. It's super important, and you can't do any of the things I'm talking about without that admin being done. But it doesn't necessarily need to be done by humans anymore. AI can listen to the whole conversation, take the transcript, and even perform actions in the background. There's nothing worse than when you speak to an agent and they say, "I'll process things one, two, three, four, and five for you at the end of this call" — and then you call back a week later because thing four wasn't done because it got lost in the chaos. If you can have AI listening in real time, transcribing in real time, keeping notes in real time, and escalating issues in real time in the background, you can also get AI to start processing the things your agent promised the customer — which takes away from the more boring parts of the job and creates some time for your agent to actually decompress afterward, rather than reliving the conversation while typing it all out.

There are so many ways this can help. Having AI that can triage conversations before they reach your agent — is this an emotionally charged person? Did they swear at the IVR coming through? Are they red hot? Let your agent know: this person is upset, be ready.

Or even having AI talk to customers coming through the IVR: "I can sense that you're really upset. I think this is a conversation for a real human being. However, I'd like to remind you that they are a real human being who's just doing their best. Please don't abuse them — they want to help you." I also think having AI summarize the customer's issues back to them is really helpful too, for that sense of: I've been heard, this conversation was useful, and it's being passed to somebody who can help me.

And then the obvious part: if it is something simple — I'm just canceling my credit card because I left it at the pub last night — I'd love for AI to handle that, because I don't need to talk to a human. It's a waste of my time, it's a waste of their time, I can click a couple of buttons and it's done. But AI has to get it right. There's nothing worse than being in a flow where you're thinking, this isn't what I want, and it's making it difficult for me to get to a human — not because it's solving my issue, but for the sake of cost cutting for your business.

Lauren Gold: That leads me perfectly to our last question. For all of us CX leaders out here trying to get executive buy-in to fix this — what's one sentence that connects emotional labor directly to business outcomes?

Jess Jackson: I actually think they're not mutually exclusive — they're mutually inclusive. If you're using AI correctly, and not even just AI — insert any cost-cutting measure, insert any business case — if you're using it correctly, it is actually beneficial for your customer, for your agent, and for your business. If it's not hitting all three, then you just shouldn't be doing it. You should be reworking that.

And frankly, as the exec we're talking to in this conversation — that's what they pay you for. That's the strategization. Not just taking the most obvious thing that will make your board happy right now.

I think every time we say we need to do "more with less" — a phrase I think everybody has heard ad nauseum at this point — we should be forced to answer: less what? Is it less time? Is it less support? Is it less humanity? Are you willing to bleed your customers over this? Are you willing to have churn and attrition of your agents over this?

If "more with less" is going to be the new normal, and if layoffs are going to be the new normal and everything else we're seeing in the industry right now, then building emotionally sustainable CX organizations has also got to be the new leadership normal. The companies that will come out of this the strongest will actually be the ones that treated emotional capacity as a strategic asset — especially during cuts — not as something to be ignored or a free, infinite resource, which it has never been.

Lauren Gold: Thank you so much, Jess Jackson. Always so insightful, always gives me a lot to think about. Really appreciate you sharing your expertise on this important topic today.

Jess Jackson: Of course, thanks for having me. This was awesome.