Customer Service at an Unprecedented Scale | Deloitte Netherlnads


Customer Service at an Unprecedented Scale

How conversational AI can manage the impact of COVID-19 on contact centres

COVID-19’s unprecedented spread has placed immense pressure on organisations. The need for clarity has caused a huge spike in calls which is dramatically increasing the burden on contact centres in all sectors. How can organisations ensure they are managing customer conversations effectively: delivering timely, accurate, and personalised information at scale? Conversational AI might provide the answer.

Dutch version

COVID-19’s unprecedented spread has put immense pressure on institutions and organisations on how to respond. Measures put in place to slow the spread have caused confusion for people, both as private citizens and consumers. The need for clarity has caused a huge spike in calls which contact centres of both public and private institutions are struggling to answer promptly: from booking COVID-19 tests to rescheduling flights; from insurance claims to cancellation clarifications. The question then arises: how can organisations respond to COVID-19, and indeed to times of crisis more generally, to ensure clear, consistent, accurate, and decision-useful information is delivered to customers or citizens in a timely fashion?


Chatbots have proven effective at deflecting much of the surge in call volume to contact centres around the globe. For example, data from shows that conversations across their clients’ existing chatbots have increased up to 250% in some industries since COVID-19 was declared a global pandemic. Other organisations have rapidly implemented COVID-19-specific chatbots to answer questions from concerned citizens. Microsoft’s Healthcare Bot has been deployed by a number of organisations in the United States, and is answering more than 1 million COVID-19 queries every day.

The reasons for utilising the technology are clear:

  • It ensures consistency across all customer interactions;
  • It delivers information immediately, any time of day or night, and; 
  • It allows for immediate scalability to meet spikes in demand.

This allows organisations to manage unprecedented spikes of queries in a consistent and timely fashion.


However, even during times of crisis, it is prudent to incorporate key best practices in the design and implementation of Conversational AI solutions. To ensure your chatbot provides immediate value and clarity to customers, it is important to:

  • Be Timely 

Time to market is a metric often associated with conversational AI solutions. Setting up the structures and governance to move and react with speed in an agile manner will ensure accurate and timely information can be conveyed to customers, even as the situation evolves. Some organisations have been very successful in this regard: the government of Estonia implemented a multilingual COVID-19 response chatbot in less than a week, to ensure everyone living in or visiting Estonia had accurate, up-to-date information from official sources regarding the virus.

  • Be Relevant

    Defining caller intents through the call centre is challenging, particularly during times of crisis. Frequent reporting on chatbot conversation data reveals what matters most to your customers, and helps prioritise development efforts in building new use cases or flows.
    Deloitte’s own chatbot, Dot, was built to support candidates through the recruitment process. Once the solution was live, we were able to uncover further topics that candidates were interested in, but perhaps too shy to ask recruiters. By querying the conversation database and what candidates were asking Dot, we prioritised our backlog to deliver additional conversation flows that would prove most valuable for our candidates.
  • Be Empathetic

    Key to delivering a positive experience is demonstrating empathy. Understanding how the customer must be feeling and actively showing this helps build trust, reassurance, and confidence in the response. Use empathetic language and be outcome driven, proactively identify solutions to customer issues. Further, tools such as sentiment analysis can help ensure the right tone of voice and response.
  • Be Personal

    Deliver real user value by providing responses specific to the end user. By gathering information through the conversation, or integrating with third party databases, chatbots can tailor their responses to meet specific user needs. The value this delivers to end users is a step change from more generic, one-size-fits-all approach. In times of crisis, this imparts confidence in the user that the situation is in control, that they are understood, and are valued to the organisation.
  • Be Consistent

    Critical during crises is the consistency of information across different channels or sources. Conflicting information can cause significant confusion, and diminish the effectiveness of the response. Chatbots ensure consistent information delivered across all interactions, but they must also be consistent with information provided by other sources including call centre agents and FAQ pages. For this reason, alignment between functions and teams is critical. Developing a master playbook, or a Chief Crisis Officer touchpoint, will ensure consistent information is delivered in all customer interactions.
Did you find this useful?