How Belmond cut response time by nearly 20 hours with AI-driven auto-tagging

Manually categorising customer contacts slowing down your support team? What about reading lengthy client-agent exchanges to grasp what’s going on? Or switching back and forth between tools to translate emails? These were the challenges Belmond faced.

The luxury travel brand boasts a global presence that spans 26 countries and territories. Their portfolio includes 45 properties, including 29 hotels, seven iconic tourist trains, seven river cruise boats, and two safaris. A diverse portfolio and high customer expectations can come with challenges, especially when it comes to managing complex, multi-contact support inquiries. Plus, servicing an international market means that many interactions occur in different languages.

Understanding just how impactful these issues can be for a company, EdgeTier worked with Belmond to implement a set of innovative solutions that leverage the power of AI-driven auto-tagging and interaction summaries.

A quick snapshot of the impact EdgeTier has had for Belmond

Let’s explore how these features not only seamlessly integrate with Salesforce but have also provided the travel company with real results, including:

• Reduced response time
• Increased agent efficiency
• Simplified workflow
• Faster onboarding
• Targeted training

Belmond’s customer support agents rely on Salesforce to manage incoming emails and the corresponding threads. Many of the conversations they receive daily are complex, spanning over 10 or more emails with some reaching upwards of 50 within a single thread. Prior to partnering with EdgeTier, an agent had to read through ALL the previous interactions before replying, which could take a significant amount of time and made it difficult to maintain optimal efficiency.

Many of the emails the agents receive are in different languages, including Spanish, Italian, and Portuguese. Yet, the majority of their agents are native English speakers. To understand what was being said, the team members had to either immediately recognise the language used in the email and send it to a fluent agent or switch to another tool for translation.

Using multiple tools is time-consuming and slows down the response process. Additionally, some of the non-English emails submitted end up being spam, resulting in an inefficient use of resources.

With such a variety of inquiry topics coming in, categorising the emails manually was a resource-intensive task. Agents had to quickly identify the topics of incoming messages and send them to the appropriate team. There was also a risk of missing critical sales emails that were time-sensitive, which could negatively impact revenue. 

One of the most advanced parts of the collaboration is conversation summaries. For this project, EdgeTier set out to make it easier for Belmond’s agents to quickly digest long email chains. Inspired by our release of AI-powered interaction summaries in the EdgeTier platform, the team created a custom version that scans all the interactions in a thread and creates three to seven bullet points that highlight the key points.

Powered by a large language model (LLM), what the technology targets and highlights is based on the client and their agents’ unique needs. Some of the examples used for Belmond include:

• The reason for the contact
• Booking details
• Any special travel requests
• Dietary restrictions

 

The customer support interaction summaries provide agents with the most important bits of information so that they aren’t forced to spend valuable time and energy reading through multiple emails.

Regardless of the original language of the emails, the summaries also automatically appear in English, which, according to Haley Abbott, the Project Manager and Salesforce Program Manager at Belmond, “helps the team triage without having to go to translate every single email.” As a result, their agents are able to effectively gather important information without the need for using a separate translation tool.

“Case summaries have been incredibly critical for us to catch agents up quickly.”

Haley also pointed out that the summaries enable Belmond’s agents to immediately catch spam. “From a quality control point of view, it’s easier for the team to see what is happening in an email with a summary,” which cuts out the additional steps of switching tools, saving agents time. 

Prior to summaries, the collaboration process between EdgeTier and Belmond began with classification. The goal was to simplify workflow by eliminating manual classification of emails. To accomplish this objective, the EdgeTier team harnessed the power of AI to quickly scan every email that’s submitted.

“The average response time for post-booking type emails is about five hours. Previously, it was over 24 hours.”

From there, AI-driven auto-tagging labels each email based on the subject matter. This feature enables them to quickly “route inquiries to the right people.” For example, if it’s a payment-based inquiry, it’s sent to the service-focused customer support agents. Or, if the interaction is related to a potential transaction, it’s routed to agents specialising in sales support.

On top of the initial classification, the system also evaluates the email for time sensitivity. The AI scans for any mention of dates within the next two weeks. If found, EdgeTier automatically flags it as high-priority so that it jumps to the front of the queue. For instance, if a potential customer is inquiring about booking a trip in the next week but needs to verify travel document requirements first, a delayed response could result in the loss of the sale. EdgeTier makes sure that doesn’t happen by immediately tagging it and getting it in front of an agent faster.

Prior to partnering with EdgeTier, the Belmond team has spent a significant amount of time and energy building out their CRM system. To prevent disrupting their workflow with another tool, the EdgeTier team integrated both the interaction summaries and tagging process directly into Salesforce. 

Once these tailored solutions were implemented, Belmond was able to start seeing results quickly. Here’s a closer look at the real-world impact: 

By implementing auto-tagging classification and prioritisation, Belmond reduced their average first response time by nearly 20 hours. Emails that previously took over a day to address are now responded to in under five hours, ensuring faster customer resolutions. 

With multilingual summaries and automated categorisation, agents can focus on delivering quality responses rather than manually processing incoming emails. For example, instead of reading and comprehending around 1,000 for a 10-email thread, agents only need to review around 50 to 70 words in the interaction summary – a reduction of roughly 90% in text volume. As a result, tasks that once required extensive back-and-forths are now streamlined, improving overall productivity. 

Our AI-powered solutions integrate directly with Belmond’s Salesforce system to create a unified, all-in-one platform. Agents can now access categorised emails, summaries, and translation features without ever leaving their CRM system, saving them time and ensuring efficiency. 

By automatically summarising key points in English and automating tagging processes, Belmond has significantly reduced the time spent processing spam and non-urgent inquiries. This ensures that resources are directed toward resolving high-priority or revenue-generating tasks first.

“[EdgeTier] came to where we were instead of the other way around, which we don’t always get with other vendors.”

New agents can get up to speed quickly with the help of categorisation and summarised conversation histories. The ability to prioritise and route less critical inquiries also provides trainees with a safe space to practice and improve. Additionally, the data generated from the initial categorisation and processing helps point out any response-time differences between support teams and office locations, which can be used to identify a need for additional training or support. 

With EdgeTier’s tagging and analytics capabilities, Belmond’s support team now has the data to help demonstrate their contributions to the organisation, such as the quickness of their first response time. This increased visibility allows for more informed decision-making and helps solidify the value of the support function within the company. 

Improving the efficiency of customer support operations begins with smart solutions, like EdgeTier. Our collaboration with Belmond highlights both the power and the flexibility of our technology to meet the unique needs of companies with high-complexity support contacts.

It also highlights our team’s commitment to a collaborative mindset. According to Haley, the EdgeTier crew were “super easy to work with and understanding of our use cases. Really open-minded about finding a solution to what we’re trying to do, and never pushy.” Since day one, we’ve been dedicated to developing innovative solutions that match a customer’s needs, not the other way around.

Looking to reduce your response times, improve agent efficiency, and simplify your support operations? Let us show you how AI-driven auto-tagging and interaction summaries can integrate directly into your CRM.