Leading Restaurant & Retail Chain cuts cost per contact by 67% and achieves 92% resolution rate with agentic AI

Before Quiq, the guest relations team was working with tooling that could not keep pace with growth, and the consequences compounded quickly:

• No live channels at all. The backlog got bad enough that the team shut off phones and chat completely rather than let response times get worse. As the VP put it: “We had no phones turned on. We were not doing any live conversations at all.”
• A seven-month response lag. By the time the team re-engaged, they were roughly seven months behind on guest responses from the prior holiday season, with 30,000 tickets unresolved.
• Holiday peaks made everything worse. Overreliance on web-based forms and email meant that every seasonal spike in volume landed directly on the same understaffed queue, with no automated way to triage or resolve simple requests.
• Unsustainable cost per contact. At roughly $12 per contact, the previous approach could not scale with the volume growth the loyalty program was driving.
• A previous vendor that didn’t solve it. The team had already brought on another tool before Quiq, but it wasn’t enough to resolve the backlog or restore live support.

Quiq built the company’s first AI agent, trained on its own knowledge base content, to answer guest questions in natural language rather than routing everything to a human queue. The system was designed to:

• Resolve routine questions without a human agent. The AI agent handles common guest inquiries directly, pulling from the company’s own FAQ and knowledge content instead of a fixed script.
• Classify every conversation automatically. Using LLM-based topic classification, the AI agent tags each closed conversation into categories such as store information, retail products, gift cards, and customer service, feeding that data straight into reporting.
• Meet guests on the channels they prefer. The company opened web chat and a messaging channel (Apple Messages for Business) alongside phone, rather than funneling everyone into the same overwhelmed queue.
• Give the team real-time visibility. Integration with the company’s Salesforce environment gave guest relations leadership a live view of queue volume and guest sentiment, rather than relying on after-the-fact reporting.
• Launch fast. The AI agent went live across the company’s website in about six weeks, including the internal testing and change management needed to get the guest relations team comfortable with the new workflow.

For the guest relations team, the AI agent changed what the job looked like day to day. Routine questions that used to consume the whole team’s time now resolve without a human agent, and escalation rates dropped noticeably as the team headed into its next holiday peak.

That freed-up capacity didn’t just reduce workload, it expanded what the team could take on. The company began preparing to open live chat directly to its field teams and take on social response management, work that wasn’t possible while the team was buried in backlog.

According to the VP of Loyalty and Digital Experience, the shift also changed morale: “The team learned that technology can help them do a better job. The AI agent has taken a lot of the tasks off their plates that they don’t enjoy doing, and has freed them up to do other work. And they didn’t know it at the time, but they love this other work.” She added that turnover on the team dropped to zero after launch.

By replacing a stalled, backlog-choked support model with a self-service AI agent, the company turned a seven-month crisis into a same-day response operation, without adding headcount.

Key outcomes:

• 92% resolution rate for AI-assisted guest inquiries
• 67% reduction in cost per contact, from roughly $12 down to $3–4
• 30,000-ticket backlog cleared, eliminating a seven-month response lag
• 10,000 chats handled in a single day during a holiday peak, equal to two typical weeks of volume, with declining escalation rates as the holiday approached
• Zero turnover on the guest relations team since launch, per team leadership
• Expanded scope for the team, including plans to open live chat to internal field teams and take on social response management