From Chatbot Builder to AI Customer Service Agent
Tidio launched in 2013 as a live chat platform and grew into one of the most popular SMB customer service tools by offering an accessible, ecommerce-native live chat solution. The introduction of Lyro in 2023 represented a significant strategic shift — from a tool that enables human agents to a platform where an AI agent handles the majority of customer interactions autonomously.
Lyro is powered by Claude (Anthropic) in its backend, which explains the natural conversational quality of its responses. Unlike previous-generation chatbots that required extensive intent mapping and flow building, Lyro simply reads your existing content and answers questions based on what it learns. This fundamental architecture change is what enables non-technical teams to deploy serious AI customer service in hours.
How Lyro Training Works
Setting up Lyro begins with feeding it knowledge from three sources: your help center URL (which Lyro crawls automatically), uploaded documents (PDFs, CSVs, Word documents), or a direct website URL. Lyro processes this content, indexes the key information, and begins answering questions based on what it finds.
The quality of Lyro's responses is directly proportional to the quality and comprehensiveness of the knowledge base it's trained on. Companies with well-maintained help centers, clear return policies, detailed product documentation, and up-to-date FAQs will achieve the highest automation rates. Those with sparse or poorly organised documentation will find Lyro frequently escalating to human agents due to insufficient knowledge context.
Ongoing training happens through the Tidio dashboard, where teams can review conversations where Lyro struggled, add new content to address knowledge gaps, and refine responses to specific query types. This feedback loop is straightforward but requires regular maintenance to keep Lyro's knowledge current as products and policies evolve.
The 67% Automation Rate: What Does It Actually Mean?
Tidio prominently advertises that Lyro can handle up to 67% of customer queries autonomously. This figure deserves context. In ideal conditions — ecommerce businesses with comprehensive help centers, clear return policies, detailed product information, and primarily transactional customer queries — automation rates in the 60-70% range are genuinely achievable.
In our evaluation, we found that businesses with well-structured knowledge bases and primarily FAQ-type queries (shipping times, return policies, order status, product specifications) achieved 55-65% automation. Businesses with complex technical products, multi-step troubleshooting scenarios, or sparse documentation achieved 30-45% automation. Setting realistic expectations based on your specific query mix is important for evaluating ROI.
Ecommerce Platform Integration
Lyro's ecommerce integrations are among its strongest differentiators. The Shopify integration allows Lyro to check real-time order status, retrieve shipping information, access customer purchase history, and in some cases initiate return processes — all within the AI conversation. This moves Lyro beyond a pure FAQ bot into an agent that can resolve transactional queries end-to-end.
WooCommerce, BigCommerce, Magento, and PrestaShop integrations offer similar (though slightly less deep) transactional capabilities. For D2C and ecommerce brands, this operational integration is what justifies Lyro over simpler FAQ chatbots — it's not just answering questions, it's completing support actions.
Multichannel Unified Inbox
Beyond the AI automation, Tidio provides a unified inbox for all customer conversations across live website chat, email, Facebook Messenger, Instagram DMs, and WhatsApp (Plus plan). Human agents manage all channels from a single interface, with Lyro handling automation across each channel simultaneously. This multichannel approach is particularly valuable for ecommerce brands with active social media customer service operations.
Analytics and Performance Monitoring
Tidio's analytics dashboard shows Lyro automation rates, average response times, customer satisfaction scores (where collected), query category breakdowns, and escalation rates. For managers evaluating AI customer service performance, these metrics provide the visibility needed to continuously optimise Lyro's knowledge base and identify where human agents are still most needed.
The conversation quality review tools allow managers to read through AI conversations, flag problematic responses, and use these as training data to improve future performance. This human-in-the-loop quality control loop is essential for maintaining Lyro quality over time, particularly as products, policies, and customer query patterns evolve.
