Key Takeaways
- 96% of consumers say empathy from support agents is important, but only 52% report actually feeling treated with empathy when they contact customer support
- Explicit emotional validation from agents lifts inferred CSAT scores by 12% on average, and up to 14.8% in tech and SaaS environments, based on analysis of 35.1 million conversations
- Only 18% of customer support interactions include any form of explicit emotional acknowledgment from the agent, meaning the majority of calls leave the expectation gap unaddressed
- Emotionally connected customers are 52% more valuable than those who are merely satisfied, spending more across their lifetime and churning at lower rates
- 64% of contact center organizations are not prioritizing emotional intelligence training despite empathy being consistently cited as the most lacking agent skill in QA audits
Customer support agent empathy has more data behind it than most contact center leaders realize, and the numbers are uncomfortable.
Ninety-six percent of consumers say empathy from customer service agents is important during a support interaction. Only 52% report actually feeling it. That 44-point gap is not a communication strategy problem. It is an execution problem, and it shows up in CSAT scores, churn rates, and agent retention data in ways that carry real dollar consequences.
The analysis below pulls from Genesys, Level AI, Qualtrics, Forrester, Zendesk, Harvard Business Review, and McKinsey research to show where the gap exists, what it costs, and what contact centers that close it are actually doing differently.
The empathy gap: what customers expect versus what they get
The Genesys global consumer survey is one of the most cited data points in contact center research: only 52% of consumers feel they are shown empathy when contacting customer support. Forrester's research puts the number lower, with only 34% of customers reporting they felt consistently treated with empathy across their service interactions.
Neither figure is a rounding error. They represent the majority of customer interactions failing on a dimension that customers themselves have ranked as more important than speed of response.
Five9 surveyed consumers on what defines excellent customer experience. Eighty-six percent said empathy and human connection matter more than response time. The conclusion is not that speed is irrelevant, but that speed without acknowledgment lands poorly. A fast interaction that leaves a customer feeling dismissed tends to generate worse satisfaction scores than a slower one where the agent made them feel heard.
Qualtrics' 2026 Consumer Experience Trends report, which surveyed over 20,000 consumers across 14 countries, found that good customer service drives satisfaction for 92% of respondents. The same report found that 1 in 5 consumers who used AI for customer service received no benefit at all, a failure rate nearly four times higher than AI use in other contexts. The organizations seeing that failure rate are generally the ones that deployed AI primarily to cut costs rather than to improve experience.
How empathy moves CSAT scores
Level AI analyzed 35.1 million conversations across 85 companies and produced some of the sharpest empathy-to-CSAT data available. When agents explicitly validated customer emotions during an interaction, conversations scored an inferred CSAT of 3.94 out of 5. Conversations without that validation scored 3.52. That gap holds across industries and company sizes.
The lift is not uniform across sectors:
| Industry | CSAT lift from explicit emotional validation |
|---|---|
| Technology and SaaS | +14.8% |
| Financial services and insurance | +14.3% |
| Retail and e-commerce | +11.2% |
| Healthcare services | +13.1% |
| Telecommunications | +10.8% |
Source: Level AI, analysis of 35.1 million conversations via CX Dive
What the Level AI data also found: only 18% of conversations included any explicit emotional acknowledgment from the agent. Fewer than one in five interactions. The empathy gap is not a fringe problem concentrated in a handful of underperforming teams. It is the default state.
SQM Group's QA benchmarking programs track empathy as one of 22 scored CX dimensions alongside process adherence, compliance, and resolution accuracy. Contact centers targeting composite QA scores of 85 to 95% consistently find empathy scores pulling the average down when agents are not specifically coached on it. The metric is tracked, it shows up in audits, and it still fails to register in the majority of recorded interactions.
Empathy and customer retention
The business case for agent empathy is not primarily about CSAT scores. It runs through retention.
Harvard Business Review research on emotional connection found that emotionally connected customers are 52% more valuable than those who are merely satisfied. Fully connected customers spend 23 to 100% more than disconnected ones depending on product category, and they are significantly less price-sensitive. The distinction between "satisfied" and "emotionally connected" is made in the support interaction as much as in the product itself.
Gartner's research found that organizations in the top decile of their industry's Global Empathy Index sustained a 10% increase in customer loyalty compared to peers. That compares to the pattern in Forrester's 2025 Global Customer Experience Index, where 21% of brands declined in CX quality, 73% stayed flat, and only 6% improved. The majority of companies are not moving the needle, and the ones that are tend to have invested in the human-interaction layer rather than trying to automate their way to better experience scores.
Zendesk's 2025 CX Trends Report found 63% of consumers are willing to switch to a competitor after a single bad experience, a figure that grew 9% year over year. McKinsey documented one case study where CX investment centered on emotional understanding cut churn by 75% and drove revenue growth three times faster than competitors over a three-year window.
The churn math matters for scale: a 5% decrease in customer churn can boost revenue by 25 to 95% depending on industry, according to multiple retention studies. The sensitivity is high enough that empathy-related improvements in FCR and CSAT translate into meaningful top-line impact when measured across a full customer base.
Qualtrics estimates that $3.7 trillion in global sales are at risk annually from poor customer experiences. US organizations account for $856 billion of that figure. Lack of empathy in service interactions is consistently cited as a primary driver.
The AI empathy problem
The empathy question has a specific shape in the context of AI-assisted and fully automated customer support.
SurveyMonkey's customer service statistics found 79% of Americans strongly prefer interacting with a human agent over an AI agent, with the preference even more pronounced among older demographics. Among Gen X, only 7% preferred AI. Among boomers, 4%.
Zendesk's 2025 CX Trends data adds nuance: 64% of consumers say they are more likely to trust AI-driven customer service if it exhibits human-like traits such as friendliness and empathy. The implication is that the rejection of AI is not categorical. It is a rejection of AI that feels cold or transactional.
Hiver's 2025 AI vs. Human in Customer Service report found 52% of customer service professionals observed that customers often prefer human agents specifically for empathy and understanding. Fifty-nine percent of support professionals said they believe in a human-led support strategy for complex, high-stakes issues, even in teams that have deployed AI broadly.
The Qualtrics 2026 finding on AI failure rates points at something concrete: when AI is deployed to process volume rather than to handle experience, the empathy gap widens. Customers in those interactions frequently report feeling like they were managed rather than helped.
Genesys took a different approach, deploying conversation AI to detect empathy displays rather than replace human interaction. Their system identified more than 700 million instances of agent empathy across consumer interactions in fiscal year 2025. The data was used for coaching and QA scoring rather than automation, which represents a more defensible use of AI in empathy-sensitive environments.
Empathy training: what works and what the numbers say
Sixty-four percent of organizations are not prioritizing emotional intelligence training, according to contact center training research via Insight7, despite empathy being named as the most lacking agent skill in QA audits.
The disconnect between what QA finds and what gets addressed in training is one of the more puzzling patterns in contact center operations. The data exists. The gap is measured. And most organizations still do not run structured empathy programs.
What happens when they do:
Contact centers that build structured empathy culture report fewer escalations, lower burnout, and higher agent retention. The Zendesk 2025 CX Trends data found that companies leading in human-centric service report 22% higher customer retention rates, 33% higher customer acquisition rates, and 49% higher cross-sell revenue versus industry peers. That sample covers 10,000 or more respondents across 22 countries.
Agents trained in empathy report feeling better prepared for difficult calls. That preparation correlates with lower burnout and higher job satisfaction, both of which matter in an industry running 30 to 45% annual attrition rates. Customer support agent burnout statistics show burned-out agents score 15 to 25% lower on quality monitoring. Training that reduces the emotional exhaustion of difficult interactions helps the agent and the quality numbers simultaneously.
The connection to first contact resolution is also direct. A 15% improvement in FCR produces a 57% decrease in repeat call volume. Empathetic handling improves FCR because customers who feel heard tend not to call back to relitigate the same interaction. They either accept the resolution or escalate once, rather than cycling through multiple contacts trying to get someone to acknowledge the problem properly.
What empathy looks like in quality assurance
Most contact center QA programs now score empathy as a distinct dimension rather than folding it into general communication quality. SQM Group's benchmarking framework tracks it explicitly across 22 CX metrics. Balto and similar conversation intelligence platforms flag empathy-language patterns in real time.
The shift toward 100% interaction scoring via AI means QA teams can now track empathy rates at scale instead of sampling 1 to 5% of calls. When Level AI applied this approach across 85 companies and 35.1 million conversations, they found that 18% empathy-expression rate. That number is now measurable, auditable, and reportable in ways it was not four years ago.
Three behaviors consistently appear in high-empathy QA scores:
- Direct acknowledgment of the customer's stated emotion or frustration before attempting resolution ("That sounds incredibly frustrating, let me look at this right now")
- Personalization that references specific details from the customer's account or situation rather than generic sympathy
- Language that keeps the agent as the subject rather than the system ("I want to make sure this is right for you" rather than "Our policy states...")
Contact centers that coach specifically on these behaviors and track them in QA see 3 to 5 CSAT point improvements within two to three months of consistent coaching, according to CustomerServiceManager.com benchmarks. Five CSAT points in a competitive vertical has retention and referral value that shows up in revenue.
Empathy across customer support channels
Channel affects empathy expression and perception differently.
Voice interactions give agents the most empathy tools available: tone, pacing, silence, and the full range of vocal affect. They also carry the highest emotional stakes. Customers on voice calls tend to be dealing with more complex or frustrating issues than those in chat or email.
Chat introduces constraints. An agent running three concurrent sessions cannot provide the same deliberate, single-focus empathy as a voice agent. The words are available, but the timing and personalization suffer under concurrency pressure. Customer support average handle time statistics data shows that AHT pressure has a direct empathy cost: agents optimizing for short interactions produce shorter empathy windows.
Email gives agents time to craft language carefully, but loses the real-time acknowledgment that makes empathy feel responsive rather than scripted. A thoughtful email response 24 hours after a frustrating experience reads differently than immediate acknowledgment on voice.
Social media channels add public visibility pressure. Agents handling public-facing complaints must balance empathy with the reality that their response is visible to observers, which changes tone in ways that can read as performative rather than genuine.
The channel with the best natural fit for empathy training ROI is voice, because it allows the most variables to be coached, measured, and repeated. Chat is second. Both can be tracked at scale with current conversation intelligence tools.
How outsourced teams approach agent empathy
Well-run customer support outsourcing operations generally run more structured empathy coaching programs than in-house contact centers at comparable scale, because they have no choice. Delivering measurable CSAT improvement is part of what the client relationship requires.
Established BPO providers tend to run:
- Empathy scoring as a weighted QA dimension from the first month of onboarding
- Regular call calibration sessions where QA analysts and agents review specific empathy moments together
- Side-by-side coaching specifically targeting the 18% empathy expression gap
- Real-time dashboards that flag interactions where emotional validation was absent
Virtual assistant services that handle customer support functions use similar frameworks. The difference from traditional BPO is typically the staffing model rather than the training approach.
ICMI benchmarks show outsourced operations posting customer satisfaction scores 10 to 15 percentage points higher than in-house centers of comparable size in comparable verticals. Some of that reflects workforce management infrastructure. A portion reflects deliberate empathy training investment that in-house teams often deprioritize when resources are constrained.
Key numbers at a glance
| Metric | Value | Source |
|---|---|---|
| Consumers who say empathy is important | 96% | Dixa |
| Consumers who feel treated with empathy | 52% | Genesys |
| Consumers who feel consistently treated with empathy | 34% | Forrester |
| Interactions with explicit emotional validation | 18% | Level AI |
| Average CSAT lift from emotional validation | 12% | Level AI |
| CSAT lift in tech/SaaS from emotional validation | 14.8% | Level AI |
| Emotionally connected vs. satisfied customer value | 52% more valuable | HBR |
| Consumers willing to switch after one bad experience | 63% | Zendesk 2025 CX Trends |
| AI users receiving no benefit from AI support | 1 in 5 | Qualtrics 2026 |
| Preference for human over AI agent | 79% | SurveyMonkey |
| Orgs not prioritizing empathy training | 64% | Insight7 / training research |
| Revenue at risk globally from poor CX | $3.7 trillion | Qualtrics / Forbes |
| CX leaders: customer retention advantage | +22% | Zendesk 2025 CX Trends |
What the data points to
The empathy gap in customer support is not a mystery. It is a documented, measurable pattern: customers expect acknowledgment, most interactions skip it, and the revenue consequences are traceable through retention, CSAT, and lifetime value data.
The 18% explicit-empathy rate from Level AI is the number that deserves more attention in contact center operations. Not because it is surprising, but because it is specific enough to be actionable. A team at 18% can set a target of 35%, coach toward it, and measure whether QA scores and CSAT move. They almost always do.
The organizations that close the gap consistently share a few features: QA programs that score empathy as a weighted dimension, coaching sessions that replay specific moments rather than general feedback, and staffing ratios that give agents enough time between interactions to actually engage with each call rather than surviving it.
The AI question adds a layer. Automation handles volume. It does not handle the 86% of customers who say human connection matters more than speed. The contact centers that are threading this correctly are using AI for operational efficiency and human agents for emotional accuracy, rather than treating them as direct substitutes.
Related data: customer support agent burnout statistics, customer support quality assurance statistics, agent coaching statistics, agent engagement statistics.
Data in this article draws from Genesys Consumer Survey on Empathy in Customer Support, Level AI analysis of 35.1 million conversations via CX Dive, Qualtrics 2026 Consumer Experience Trends Report (20,000+ consumers, 14 countries), Zendesk 2025 CX Trends Report (10,000+ respondents, 22 countries), Forrester 2025 Global Customer Experience Index, Five9 consumer survey on empathy vs. speed, Dixa customer service empathy research, Harvard Business Review emotional connection research, McKinsey experience-led growth research, Gartner Global Empathy Index analysis, Hiver 2025 AI vs. Human in Customer Service Report, SurveyMonkey customer service statistics 2026, SQM Group QA benchmarking research, CustomerServiceManager.com empathy training benchmarks, Insight7 contact center training research, and Genesys FY2025 empathy detection announcement.
Frequently Asked Questions
What percentage of customers value empathy in customer support?
96% of consumers say empathy from customer service agents is important during a support interaction, according to Dixa research. Separately, 86% of consumers say empathy and human connection matter more than speed of response in delivering excellent customer experience (Five9 survey). Despite this, only 52% of consumers report actually feeling treated with empathy when contacting support (Genesys).
How much does agent empathy improve CSAT scores?
Level AI's analysis of 35.1 million conversations found that explicit emotional validation from agents lifts inferred CSAT scores by 12% on average. The lift reaches 14.8% in technology and SaaS environments and 14.3% in financial services and insurance. Contact centers that coach specifically on empathy behaviors see 3 to 5 CSAT point improvements within two to three months.
What percentage of customer support interactions include empathy?
Only 18% of customer support conversations include any form of explicit emotional acknowledgment from the agent, according to Level AI's large-scale conversation analysis. This means fewer than 1 in 5 interactions address the empathy expectation that 96% of customers say matters to them.
Does empathy training work for contact center agents?
Contact centers that run structured empathy training programs report fewer escalations, lower agent burnout, and higher retention. Zendesk's 2025 CX Trends Report found that companies leading in human-centric customer service post 22% higher customer retention rates, 33% higher acquisition rates, and 49% higher cross-sell revenue compared to peers. Agents trained in empathy also report higher job satisfaction, which helps in an industry running 30 to 45% annual attrition.
Tags
Ready to put this into practice?
Book a free 15-min match call
Tell us what role you're filling. We'll match you with a pre-vetted virtual assistant - or tell you honestly if we're not the right fit.
Book a free call →