Logo Reinfer

Hiscox Services: Enhancing CX and operational excellence with NLP

Hiscox Services: Enhancing CX and operational excellence with NLP
Hiscox is a global leader in specialist insurance, covering both people and businesses. With over 100 years’ experience and 3,000 employees, Hiscox delivers fine-tuned customer service and coverage that meets the needs of today.
Customer since
“Re:infer has completely changed the way we see service. It used to take three weekdays for a broker to receive a response from our team. Now it takes less than three hours. Our brokers understand demand and value-add in a completely different way.”
<small>Marco Rodrigues, Head of Service Centre, Efficiency and Automation Manager at Hiscox<small>

The challenges

As a major specialist insurance provider, Hiscox aims to deliver the best protection through high-quality products, excellent service and claims support. The largest part of its business is property and speciality commercial insurance for businesses. It also provides speciality insurance products for high-net-worth individuals with considerable assets and capital. Competition for these elite customer cohorts is fierce and they are used to receiving exceptional service that’s responsive and tailored to their needs.

Hiscox deals with two primary inbound service channels: brokers represent around 80% of its business, with the rest composed of direct customers. In the UK alone, it does business with a network of more than 2,000 brokers. Communication between Hiscox underwriters, brokers and customers is critical to success.

Hiscox prides itself on its claims and broker services, and strives to achieve the gold standard for the industry. However, huge volumes of inbound emails were causing a lack of visibility in one of Hiscox’s most crucial service channels. While brokers can engage with Hiscox via portals and calls, the primary driver of this demand was email. Underwriters were forced to process every broker email manually, creating bottlenecks at key points of the communications process that impacted service speed and degraded the customer experience.

Under this model, there was zero visibility or control into service metrics, and task allocation based on skills or product expertise wasn’t possible. There was no clear way to expand capacity without increasing headcount and costs, while also bringing more inefficiency to the service provided.

“Understanding means more than ‘having an idea’ about demand. Understanding, for us, means mapping everything we are receiving and trying to allocate it to the right people, creating a consistent workflow. We are achieving this with NLP and Re:Infer.”
<small>Marco Rodrigues, Head of Service Centre, Efficiency and Automation Manager at Hiscox<small>

The solution

As Hiscox continued to grow its specialist insurance business, service leaders recognised they would have to adapt their operational processes to scale with rising demand. Hiscox realised it needed to increase service capacity without unsustainably increasing overheads and complexity.

That’s why Hiscox Services partnered with Re:infer to create a new tech-enabled operating model for its underwriting and service delivery teams. The Re:infer Conversational Data Intelligence platform has provided an extra layer of intelligence to the service process, understanding and interpreting every email communication at speed and scale.

This new service model was achieved over three distinct phases:

Phase 1: NLP for complete service intelligence

Hiscox’s underwriters had limited visibility across their workflows and struggled to report on the type demand they processed on a daily basis. Precious time was being wasted on repetitive manual comms tasks, like triaging emails to the right team after being sent to the wrong person.

The first objective was to gain an understanding of the type and quantity of inbound demand agents received. Re:infer was deployed to service mailboxes, creating a first line of understanding and intelligence. The platform immediately extracted the important information from inbound emails as structured data. This data was then sent to an automated routing solution able to allocate each request to the best people for the job.

Hiscox now had an efficient, intelligent workflow management system for the categorisation of work. Instead of every underwriter being responsible for sorting inbound demand, people were left to focus on the right task for their specialist skills.

Phase 2: End-to-end automation

The structured data created by Re:infer has also allowed Hiscox to deploy end-to-end automations in its service function. Using pre-built connectors for UIPath RPA, Re:infer allows the insurer to understand and automate from end to end its transactional service requests.

The structured data created by Re:infer NLP is fed directly downstream to RPA. Around 5% of all service requests have been automated with zero human touch, and a further 80% are in scope for automation down the line.

Phase 3: The road to deep automation

Hiscox is now laying the foundations for deep automation across its service function. The end goal is to combine NLP, RPA and OCR technologies to read both emails and their attachments, before these requests are processed by downstream automation. This will greatly increase the scope of automatable requests, enabling Hiscox’s skilled underwriters to focus on their work.

Re:infer's auto-routing deployment at Hiscox Services.

The results

Hiscox Services has gained total visibility over its email demand, and the structured data needed to deploy automation and allocate its people where they’ll have the greatest impact. With Re:infer, Hiscox can operate this service model consistently and reliably, ensuring brokers and customers receive a quote or solution in good time - every time.

Re:infer has also enabled unprecedented upskilling across the service team. Hiscox Services has been able to take admin-focussed, back-office agents and turn them into NLP model specialists with Re:infer. Thanks to Re:infer’s no-code interface, non-technical agents are able to directly improve the service experience by training Re:infer models.

Re:infer is currently being rolled out to Hiscox Europe and Hiscox USA's service teams, and will soon be deployed into the vital claims function as well. Another Re:Infer use case is being explored for Hiscox’s sales channels and customer contact call centre to better understand customer demand and purchasing drivers.

Key results

  • Reduction of 300% in process lead time: from 3 weekdays to under 3 hours.
  • 5% of all processes initially automated E2E, with another 80% of processes in scope for automation.
  • Over 95% accuracy in auto-triage and work allocation.
  • 50 FTE saved in one year alone compared to manual sorting and triage.
  • £370k back to the business in Year One
“At Hiscox, we have a motto for how we want to do service: ‘easy, simple and worry-free’. I like to use this to describe Re:infer - easy, simple and worry-free technology. If I want to launch a new model I don’t need to ask for a new use case or budget, I just ask my team to start model training.”
<small>Marco Rodrigues, Head of Service Centre, Efficiency and Automation Manager at Hiscox<small>

Watch the Hiscox Services Story

Hiscox’s Head of Service Centre, Marco Rodrigues, joined SSOW 2022 - Europe’s largest European shared services event - to discuss how Re:infer is helping his team transform service quality and efficiency. Watch his full presentation:

Turn any conversation into actionable data

Drive operational efficiency and optimise the customer experience with Re:infer’s Conversational Data Intelligence Platform.