From Friction to Seamless: How Secret Escapes Transformed Their Contact Center with Natterbox AI
Secret Escapes is a leading members-only travel company offering exclusive, hand-picked luxury hotel stays and holidays at insider prices. Operating across UK and German markets, the business has always positioned itself on the quality of its curated experiences, ranging from hotel-only stays to fully packaged holidays including transfers, excursions, and flights. Founded in 2011 and with a contact center team managing thousands of inbound contacts each month, maintaining premium service while achieving operational scale has become a defining challenge.
30%
of all calls handled by AI Agents
6 weeks
From contract to AI live in production
CSAT
Scores maintained post-implementation
Challenge
Secret Escapes’ contact center faced a familiar tension in fast-growing businesses: the company sells high-value, emotionally significant holidays, yet the systems supporting its customer service had not kept pace with the volume and complexity of inbound demand.
A legacy IVR
What began as a simple menu-based IVR had evolved over time as the business grew. The team kept adding to it: more options, more self-service content, more branches, all in an attempt to triage calls more effectively.
“We almost created a monster out of it. We thought we were being tactical in routing the calls, but in fact we were creating frustration at the beginning of the call.” – Ash Bhimji, Head of Customer Services & Travel Admin
The unintended consequence was that members were spending longer than necessary navigating menus before reaching an agent. Those who selected the wrong option arrived at the wrong team and had to repeat the reason for their call from scratch. It added friction, increased handle time, and put pressure on CSAT scores.
Scaling without spiralling costs
With a contact center team that had grown to around 80 agents, the business faced a structural problem: as inbound contacts increased, the cost of handling them grew at the same pace. The question, as Ash frames it, was “how do we scale effectively without having to increase our spend year on year?” That question led the business toward AI.
A previous AI project that never delivered
Secret Escapes had attempted to deploy conversational AI before. A competing vendor had been engaged on a project that stalled after over a year without going live – leaving the team sceptical of AI promises and cautious about committing to another implementation.
Solution
Natterbox deployed a conversational AI Agent across Secret Escapes’ UK support line, targeting the two highest-impact use cases immediately: pre-sale lead qualification and general enquiries. The approach was deliberate: start where volume was highest and repetition most costly, prove value fast, and expand from there.
Conversational AI replacing the IVR
Rather than presenting callers with a menu, the Natterbox AI engages members in natural, open-ended conversation from the first interaction. The AI identifies intent, whether a caller wants to book a holiday, has a post-booking query, or needs general assistance, and routes or resolves accordingly.
The result is the end of the menu-maze. Members with straightforward enquiries are resolved without a human agent. Members who need a specialist are routed directly, with no wrong-option penalty.
Context transfer: the feature that changed everything
The single feature that most influenced Secret Escapes’ decision to choose Natterbox was the seamless AI-to-human handoff. When a call needs to escalate, the AI passes a real-time summary directly to the agent before they say hello: booking details, travel requirements, and the reason for the call.
For the Assisted Sales team, the difference is transformative. When a member calls looking to book a five-star hotel in Greece, the AI collects the destination, travel dates, party size, and specific requirements, then passes everything to the agent instantly. The agent joins the conversation already briefed, ready to search deals and provide options from the first moment of the call.
“We don’t want our customers going around a rabbit hole of AI. So it’s really important that when the customer really needs us, the agent is there on the other side. That smooth transition worked really well for us.” – Ash Bhimji, Head of Customer Services & Travel Admin
Live in six weeks
From contract to production deployment on the UK support line took six weeks. For context, a competing vendor had taken nearly a year to deliver just three to four use cases. The Natterbox team handled the technical heavy lifting, covering system prompt configuration, data endpoint setup, and integration, without requiring Secret Escapes’ internal technical teams to be involved.
“One of the things I really liked about the Natterbox implementation was that we did not need to involve our technical teams in the setup. The Natterbox team did all the heavy lifting.” Laura Meschi, AI Manager
Results
The deployment of Natterbox AI has delivered measurable operational improvement and a genuine shift in the day-to-day experience of both members and agents alike.
30% of calls handled by AI Agents
Secret Escapes set an ambitious 30% call deflection target as the primary success metric. Since implementation, handle times have been decreasing and the team is tracking consistently toward that goal. In Ash’s words: “We are on track for the 30% deflection goal by the end of the year.”
CSAT held steady, and some months beating target
The measure of success in a premium customer experience is not just cost efficiency; it is whether the member experience is protected. CSAT scores have not just held since deployment; some months they have come in above the 80% benchmark target. Handle time has been reducing, and the feedback from customers has reflected a smoother, faster experience.
“We’re a firm believer that even if handle time is increasing, our CSATs need to be maintained. So even though this is a positive story, we’re seeing handle time decrease and the CSAT has maintained. Some months it’s actually higher than our target state. That’s a positive all around.” – Ash Bhimji, Head of Customer Services & Travel Admin
A step-change in agent morale
Perhaps the most immediately felt impact has been on the team itself. The cold handoff has been eliminated: agents no longer need to ask the reason for the call, confirm booking details, or repeat questions the customer has already answered.
Laura describes the shift: “The daily morale of our human agent has been impacted positively, essentially because AI removes the frustration of a cold handoff. When a call requires a human agent, the AI produces a live summary so that when the agent gets the call, they don’t have to ask our members to repeat themselves. They get right into what they do best: create value, build a rapport, and finalise that booking.”
Ash echoes this from an operational perspective: “Our team are loving the fact that they’re getting a lot more information upfront.”
Abandonment rate falling
Customers who previously hung up during lengthy IVR navigation are now connecting with the right resource faster and staying on the line. The abandonment rate has been on a steady downward trajectory since deployment.
“It was a sigh of relief when we had our first call triaged through AI. Listening back to that call, the way the AI interacted and the way the customer interacted, it was bliss. It worked perfectly well. All my worries were put to bed. Everything’s working, and we can now ramp up steadily.” – Ash Bhimji, Head of Customer Services & Travel Admin
Looking to the Future
Secret Escapes sees Natterbox AI as a foundation to build on, not a finished deployment. The immediate next step is expanding the AI to the German support line, the business’s second-largest market, before broadening to additional use cases across voice, live chat, and email.
Laura articulates the longer-term ambition: to implement AI consistently across all contact channels so that members arriving at the Contact Us page have a seamless, coherent experience regardless of how they choose to get in touch. The goal is also to use AI to quality-control 100% of interactions through transcript analysis, something that was previously impossible at scale.
“If you are a customer experience leader and you’re hesitant, my advice would be to start simple. Just pick one simple, repetitive pain point and let the AI handle that. Try to adopt a mindset of progress over perfection. You do not need a flawless, complicated system to go live. You just need a very simple one that you can tweak with feedback and measure easily.” – Laura Meschi, AI Manager
Ready to transform your contact center with conversational AI?
