Multi-brand menswear retailer doubles self-service resolutions across 3 distinct brands.

• A chatbot that only knew fixed intents. The previous system matched customer messages to a small set of pre-written intents and returned the same static answer to everyone, regardless of which brand they were shopping, what they’d already bought, or what their actual question was. It topped out at a 25% containment rate and had nowhere left to grow without a rebuild.
• One support team, three brands, one bot that couldn’t tell them apart. Each brand has its own product lines, rental policies, and promotions. A menu-driven bot had no reliable way to route a Jos. A. Bank suiting question differently from a Men’s Wearhouse rental question, so agents ended up untangling brand-specific context by hand.
• Repetitive questions crowding out the interactions that mattered. Order status checks, rental steps, and loyalty program questions flooded the same queue as wedding-party scheduling conflicts and other situations that genuinely needed a person’s judgment.
• A hard ceiling during seasonal peaks. Prom and wedding season, along with the holiday rush, drove spikes in appointment and rental questions that the existing bot simply couldn’t absorb, pushing more volume to agents at the exact moments they had the least slack.
• Manual upkeep every time policy changed. Any update to a promotion, return policy, or store procedure meant hand-editing the bot’s fixed responses, a lag that showed up as outdated answers reaching customers.

Despite kicking off the project in the middle of peak holiday season, the team moved through three structured phases: defining scope and success metrics, a three-month build that had to account for a complex, multi-brand system architecture, and a testing and launch phase with robust QA before any customer saw the new experience.

• Quiq made it easy to get their content AI-ready. Rather than asking the team to manually reformat FAQ articles, product information, and store policies by hand, the Quiq team used the platform’s capabilities to reformat the content where necessary. The AI platform chunks that content into focused sub-articles, generates related questions, and extracts reference links automatically, and will continue to keep everything in sync with the original source documents in the future.
• Brand-aware routing built into the reasoning itself. Before answering, the AI agent identifies which of the company’s brands the customer means so it knows the brand voice and personality that it should emulate. It does this by selecting a Process Guide, a specific set of instructions and tools built for each brand as well as Guide for the use case, whether it’s locating a store, tracking an order, or capturing a sales lead. This is the piece that a fixed-intent bot structurally couldn’t do: intent-matching treats “where’s my order” the same regardless of brand, while Process Guide selection changes the policies, tone, and systems the AI agent draws on based on which brand and situation it has identified.
• Direct connections into existing systems. The AI agent works inside Salesforce OmniStudio and Manhattan Active Order Management, the same systems that already hold the company’s knowledge base, customer records, and order data, and opens a Salesforce case automatically for every conversation.
• Built-in handoff to people. When the AI agent doesn’t have the knowledge or authority to resolve something, it escalates to a human agent in Quiq’s Digital Contact Center which sits inside Salesforce, routing to the right specialist and carrying the full conversation context so the customer never repeats themselves.
• Automatic conversation scoring. An AI Analyst reviews every closed conversation, whether it ended with the AI agent or a human agent, and classifies the issue and estimated satisfaction, giving the team an aggregate view without manually sampling transcripts.
• A deliberately close partnership with IT. The VP credits weekly program management meetings and monthly business reviews with IT as a key reason the rollout held up under a complex, multi-brand system architecture: “You can’t have your head in the sand about technology, even if you’re not in IT. Being a ‘silo jumper’ and building genuine partnerships with your IT team is crucial.”

A parent texting in about a son’s homecoming suit needed to move a fitting appointment up before an order deadline. The AI agent could reschedule directly, and then answer a follow-up question about whether the order could be placed offline once the fitting was done, without asking the customer to repeat any context from earlier in the conversation.

In a separate conversation, a customer asked simply, “How do I rent a tux?” Instead of returning a generic policy blurb, the AI agent walked through the actual steps for that brand: sharing a carousel of styles to choose from, helping create an account, and then scheduling an appointment to get measured.

The support team no longer fields the bulk of routine order-status and rental-process questions, freeing them to focus on the interactions that need a person: wedding party coordination, scheduling conflicts, and other situations with real stakes attached.

That shift didn’t happen automatically. The team treats the AI agent the way they’d treat a new hire. “Everyone is responsible for the AI agent,” the VP explains. “We knew we’d need to keep an extra close eye on him in the early months and were prepared to iterate rapidly, because even though we’d QA’d him extensively, there are always real-world situations you can’t plan for, just like a human agent encounters. We maintain detailed spreadsheets tracking why something worked or what needs to be changed for better responses. The AI agent is held to the same standards as a human agent and was given grace during his ‘learning period,’ exactly like a human gets when they’re training up.”

Six months after replacing a fixed-intent chatbot with a brand-aware AI agent, the company nearly doubled how much of its support volume gets resolved without an agent, and it did so across three brands at once rather than one at a time.

• Automated resolutions rose  to 45% within 6 months and continue to climb as they iterate.
• Support coverage extended to 24/7 across all three brands, closing the gap between what the previous system could offer and when customers actually needed help.
• Knowledge updates now flow into the AI agent automatically from the source documents, removing the manual rebuild step that came with every policy or promotion change under the old chatbot.
• The team gained conversation-level observability into thousands of interactions, letting them find and close knowledge gaps directly rather than guessing at what the bot was missing.