Explore more publications!

Lyro: 28% of Shoppers Have Left a Brand Over AI, Yet Rate AI Customer Service as Highly

Resolution rates: Lyro AI 67%, Bot + human 57%, Generic bot 29%.

Resolution rates

Infographic titled "The trust gap" showing that 1 in 3 customers have left a brand over AI usage, while trust in a brand's own AI is 3x higher than in third-party agents.

The trust gap

Chart showing agentic commerce value projections for 2030: Morgan Stanley ($190-385B), McKinsey ($1T), and Ark Invest ($9T), noting that 23% of Americans purchased via AI last month.

What is agentic commerce worth?

New research draws the line between AI customer service that builds loyalty and AI deployment that drives churn — with resolution data to prove it.

Resolution rate is the eye-catcher. It is the number that tells operators whether an AI deployment is actually working, but few implementations address the perceived resolution quality, AI CSAT.”
— Bart Turczynski, Head of Growth at Tidio
SAN FRANCISCO, CA, UNITED STATES, April 1, 2026 /EINPresswire.com/ -- Research covering 22 million customer service conversations identifies the deployment gap between AI that retains customers and AI that loses them.

About 28% of consumers have already stopped purchasing from a brand because of how it uses AI, according to a global survey of more than 15,000 consumers by HubSpot and SurveyMonkey. The same research finds that 82% of consumers prefer human customer support — even when the outcome and wait time are identical to an AI-handled interaction.

Tidio's own platform data complicates that picture. Across e-commerce deployments, Lyro, the AI agent for customer service, records an average AI CSAT score of 88% of the human benchmark, with a growing cohort of businesses reaching full parity. Customers are reporting a preference for humans while consistently rating AI-handled interactions almost as highly, receiving resolutions in seconds rather than minutes, hours or days.

A new report published today by Lyro, AI in E-Commerce in 2026: The New Shopping Funnel, draws on 70 sources to examine where that gap comes from and what separates deployments that earn customer trust from those that erode it.

The operational data points to a clear pattern. Contentsquare's analysis of 22 million customer service conversations finds that hybrid human-plus-AI models resolve 57% of inquiries—nearly double the 29% achieved by bot-only deployments. The gap widens for high-stakes interactions: refund resolution drops to 18% for bot-only approaches, and technical support falls to 15%. Tidio customers deploying Lyro average a 67% automated resolution rate, reflecting the performance level that becomes achievable when knowledge coverage, escalation design, and continuous quality monitoring are built into the deployment from the start.

The churn risk from poor deployment is concrete and measurable. The HubSpot and SurveyMonkey study finds that business leaders systematically underestimate this risk, ranking loss of customer trust as their lowest AI-related concern, behind biased outputs and data security, even as 28% of consumers report having already acted on it.

"The brands losing customers are not the ones that deployed AI," said Tytus Gołas, Founder and CEO of Tidio. "They are the ones that deployed it without transparency, without clean handoffs to human agents, and without treating CSAT as a real guardrail. The brands that get those three things right are seeing resolution rates and satisfaction scores that a human-only team at scale cannot match."
"Resolution rate is the eye-catcher," said Bart Turczynski, Head of Growth at Tidio. "It is the number that tells operators whether an AI deployment is actually working, but few implementations address the perceived resolution quality, AI CSAT."

A trust asymmetry in the research adds further strategic context. Bain's analysis finds that shoppers trust a retailer's own AI agent approximately three times more than a third-party AI tool, meaning brands that deploy AI natively, within their own environment and brand voice, hold a structural advantage over those routing interactions through external platforms.
First-party conversion data from retailers that have invested in native AI supports that advantage. During Black Friday 2025, Amazon sessions involving its Rufus AI assistant that resulted in a purchase grew at five times the rate of sessions without Rufus. It is the strongest conversion signal from any AI shopping assistant reported to date.
The report includes a practical framework covering escalation design, resolution rate benchmarks, knowledge coverage audits, and quality monitoring approaches for e-commerce teams at every scale.

The full report is available for download at https://tid.io/ai_in_ecommerce_2026


Tidio is an AI-powered customer service platform that unifies live chat, chatbots, and AI agents. Its Lyro AI Agent resolves customer inquiries automatically and escalates complex cases to human operators. The platform is designed for fast-growing e-commerce businesses that treat customer service as a revenue function. https://www.tidio.com

Lyro is Tidio's AI agent for customer service, tailored to e-commerce, SaaS, and service businesses. Lyro resolves an average of 72% of incoming tickets by taking action rather than repeating FAQs, maintains an AI CSAT score approaching 90%, and doubles as an AI shopping assistant capable of increasing average order value through product recommendations and lead collection. https://www.getlyro.ai

Media inquiries: https://tid.io/newsroom

Bart Turczynski
Tidio
pr@pr-tidio.net
Visit us on social media:
LinkedIn
Instagram
Facebook
YouTube
TikTok
X
Other

Meet Lyro, Tidio's AI agent for customer service

Legal Disclaimer:

EIN Presswire provides this news content "as is" without warranty of any kind. We do not accept any responsibility or liability for the accuracy, content, images, videos, licenses, completeness, legality, or reliability of the information contained in this article. If you have any complaints or copyright issues related to this article, kindly contact the author above.

Share us

on your social networks:
AGPs

Get the latest news on this topic.

SIGN UP FOR FREE TODAY

No Thanks

By signing to this email alert, you
agree to our Terms & Conditions