Swisscom Snaps Agent Workloads In Half After Running These Two Projects

The company also reduced handling times by 25 percent and raised customer satisfaction rates

Swisscom Snaps Agent Workloads In Half After Running These Two Projects
Contact CentreInsights

Published: February 27, 2024

Charlie Mitchell

Trimming agent workloads is not a new objective for a contact center leader. Yet, many still struggle to achieve this aim in a way that’s complementary to customer experience.

Joël Viotti, Business Owner at Swisscom, and his team got this balance spot on.

Indeed, they decreased agent workloads by 50 percent after completing two major digital contact center transformation projects.

Meanwhile, mission-critical CX metrics – including customer satisfaction – reached new highs.

The two projects? Developing a customized social customer service filter and automating specific parts of customer interactions in private messages.

Project 1 – The Social Customer Service Filter

Swisscom sees it as their job to meet clients where they are. Yet, after hopping on an increasing number of social media platforms to connect with various generations, the workload has increased over the past years.

Recognizing this, Swisscom teamed up with Khoros to develop a filter that triages posts and directs messages from across these channels to the agent desktop.

In doing so, the mechanism prioritizes them considering various dimensions that suggest a level of urgency.

“Of course, that allows us to handle Big Data in a smart way by being able to respond quickly to critical customer contacts first,” added Viotti.”But, it also filters out communications the team doesn’t need to respond to immediately or at all, making space for the agents to fully focus on delivering nothing but excellent service for our customers.”

For instance, it removes comments on posts like the example below.

Yet, still, there is a chance that someone does post an urgent message on such a post.

As such, Swisscom also established a monitoring wall, which collates different channels on a single screen that supervisors keep their eyes on. That keeps a fallback mechanism in place.

Project 2 – The Expansion of Automation

Swisscom had already implemented a virtual assistant that automates defined, transactional customer journeys.

Yet, the company realized an opportunity to maximize its usage. That involved automating parts of specific conversations where live agents always ask the same questions to find a resolution.

To achieve this, Swisscom has the virtual agent gather information from customers that is relevant to their intent as they wait in the queue. That information then passes through to the live agent, who takes over the conversation with everything they need to resolve the query quickly.

“We developed the bot with our agents, and – in doing so – we set up a series of core beliefs first,” noted Viotti. “One of those beliefs was that everything the bot triggers must be relevant to the customer and their intent.

Together, we were building out journeys, developing bot flows, and cutting out additional steps that didn’t make sense for the customer.

Yet, Swisscom didn’t stop there. The team monitored the virtual agent via a bespoke analytics dashboard – again provided by Khoros.

With the dashboard, Swisscom could spot opportunities for further improvements and optimize the virtual agent to ensure high accuracy across Switzerland’s many languages and dialects.

Nevertheless, how Swisscom brought agents into the fold is perhaps the most eye-catching, as they helped uncover automation opportunities that the project team would have otherwise missed.

“Building trust with the agents by setting a clear automation vision helped them understand and embrace the purpose and benefits from which we all benefitted,” he concluded. “Everyone was buying in!”

Dear Agents, Tell Me What’s Boring!

Sharing an example of how Swisscom’s agents aided the conversational AI project, Viotti said: “On Google Business Messages, we had customers coming into the chat, writing “hello”, and that was it – not specifying a clear intent.

“We then standardized a friendly response to thank the customer for getting in touch. But, there were still many cases where we never heard back from them – and the agents reached out again before closing the case.”

Bored of this everyday, time-consuming task, the agents helped design a flow, so when the customer only says “hello”, a bot automatically responds.

Then, if the customer doesn’t reply within 24 hours, the virtual agent closes the case without an agent having to look at the case. However, if the customer does eventually send a second message, the case reopens.

“These are the smart changes I want agents to come to me with,” concluded Viotti. “I want them to tell me of cases like this where they are following processes that – ultimately – don’t make sense.”

More Lessons from the Swisscom Masterclass

Alongside cutting agent workloads by 50 percent, these two projects reduced average handling time (AHT) by 25 percent.

That’s a big win for agents, customers, and – ultimately – the broader business.

Yet, despite these excellent results – largely driven by tech – Viotti stresses that contact centers shouldn’t over-rely on artificial intelligence; human intelligence must come first.

“It’s important to have a really good understanding of the tech, but IT teams should not dominate these projects,” he said.

“I am a strong believer that technology alone is not enough. Only when humans and technology interact optimally and you engineer backward from the best customer journey possible, great experiences can be built.

It’s, therefore, important to remember that the bot is not a copy of an agent. It has strengths, and your people have strengths. Think about how you can leverage these together to make life easier for every customer – no matter if they phone in at 8am or comment on a social post at 2am.

For more insights from Viotti and his excellent team, follow him on LinkedIn.


CCaaSChatbotsConversational AITechnology Media and Telecoms

Share This Post