A Closer Look transforms shopper support from a bottleneck into a 24/7 self-serve experience.

A Closer Look (ACL) is a leading customer experience consultancy that runs mystery shopping programs for more than 300 clients across retail, restaurant, and service industries. Each client has unique shop guidelines, typically stored in PDFs with no standard format, no consistent structure, and no reliable way to stay current.

To complete their assignments, shoppers needed specific details: photo requirements, evaluation criteria, and rules unique to each client location. Finding those details inside lengthy, inconsistently formatted documents took time. When shoppers couldn’t find what they needed, they contacted the support team, and the questions were almost always the same ones.

Human agents carried the full weight of that volume. ACL’s existing chatbot relied on manual programming, which meant there was no practical way to load and maintain hundreds of unique, constantly changing client documents in that system. The data problem had to be solved before AI could even be considered.

Quiq partnered with ACL to build Ella, an AI-powered shopper support assistant. Before any AI logic was written, the data had to be fixed. ACL’s content was spread across hundreds of PDFs in inconsistent formats, some dating back years. Even small inconsistencies in capitalization or naming conventions posed real risks for an AI model trying to retrieve the right information.

Quiq used AI-powered data structuring to convert this fragmented library into AI-ready formats, applying advanced data engineering that went beyond basic cleanup. This included complex indexing and search systems to parse unstructured PDF content, and query-routing mechanisms to distinguish between a specific shop inquiry and a general FAQ lookup.

To protect accuracy at every step, Quiq implemented pre- and post-answer guardrails that verify responses before they reach a shopper. This eliminates hallucinations and keeps Ella’s answers aligned with current client guidelines, regardless of how complex or unique the client scenario is.

Ella was also built to handle real-world complexity. The AI disambiguates a shopper’s initial query, asking clarifying questions when intent is unclear, and breaks multi-part questions into sub-tasks it can handle in sequence or in parallel.

Shoppers interact with Ella through a chat interface at any hour, without waiting for an agent. When a question arrives, Ella first determines whether it relates to a specific assigned shop or a general process question, then routes its answer accordingly.

For shop-specific questions, Ella pulls from the relevant client guidelines and returns a short, direct answer. For broader questions, it draws from ACL’s general knowledge base. When intent is unclear, Ella asks for clarification before answering rather than guessing. That disambiguation step, built deliberately into the design, is what allows Ella to deliver reliable answers across hundreds of distinct client scenarios.

For complex, multi-part questions, Ella breaks the inquiry into manageable sub-tasks and handles them in sequence or in parallel, delivering thorough assistance without requiring the shopper to navigate multiple conversations or contacts.

Before Ella, support staff answered the same questions repeatedly from shoppers who needed small but critical details to complete their assignments. That time is now available for the work that genuinely requires human judgment, including complex shopper issues, client escalations, and higher-value operational tasks.

The data standardization effort also made a lasting operational difference beyond AI. With client guidelines now consistently structured and maintained in one place, tracking updates and ensuring accuracy is far simpler than managing a library of inconsistent PDFs across hundreds of clients.

• Shoppers receive instant, accurate answers at any hour without waiting for agent availability
• Human agents are no longer burdened by routine, repetitive inquiries
• Client guidelines are standardized and AI-ready, making ongoing maintenance far simpler across 300+ clients
• Ella connects to ACL’s proprietary platforms for real-time monitoring and workflow automation
• ACL has a proven AI foundation to extend into deeper automation and new platform integrations

ACL’s investment in AI delivered measurable operational improvements across the business:

• 24/7 query management: Ella provides instant, accurate responses at any hour, eliminating wait times for shoppers
• Repetitive workload eliminated: Human agents focus on complex shopper issues and strategic tasks rather than routine inquiries
• Improved workflows: Ella integrates with ACL’s proprietary platforms, enabling real-time monitoring and advanced automation
• Version control simplified: Standardized, AI-ready data replaces a library of inconsistent, hard-to-maintain PDFs spanning hundreds of clients
• Scalable AI foundation: ACL is extending Ella’s capabilities into deeper automation and new integrations, using this deployment as proof of concept for broader AI adoption across the business