The coronavirus pandemic has sparked a massive increase in contact-centre call volume from anxious customers at the very time banks and insurance companies are having to close down their contact-centre buildings and losing staff to illness or self-isolation. Many firms can’t cope, resulting in long wait times and disconnected calls.
Contact centres are struggling to cope because call volumes are skyrocketing as employees are worried about going to work. Companies with offshore contact centres in countries like India and the Philippines have been caught out by the lockdown in those countries and most other contact centre buildings either have been or will need to be closed because of the social distancing measures now in place in many countries.
This is exacerbated by many firms not being set up for remote operations. While virtual contact centre operations aren’t a new idea, the reality is that these staff at traditional banks and insurers don’t have the laptops or secure remote access to the CRM, call-routing, customer identification, knowledge management and other systems they need to support customers from home. Corporate virtual private networks (VPNs) are struggling to cope with far larger numbers of simultaneous users.
Use messaging on digital touchpoints to triage incoming pandemic-related questions.
The answer is to use digital touchpoints to help as many customers as possible. In particular, firms should use secure messaging to triage incoming questions, enabling human agents to focus on the most complex and urgent customer emergencies.
The failure to cope with unprecedented call volumes, means that many banks and insurance companies are delivering a poor customer experience at the worst time for their anxious customers. Financial services companies need to get fast-changing information and complex guidance to millions of customers quickly.
Human support is reassuring for customers who can get through but won’t scale. The only scalable solution is to use digital touchpoints, particularly websites, apps and secure messaging, to handle most customer interactions so that human support can be focused on the customers who need it most.
Many of the enquiries that are flooding into banks and insurance companies are specific questions about an individual’s situation that can’t be answered with generic information. To handle specific questions, firms need secure messaging or chat. Secure messaging offers customers an alternative way to communicate with banks and insurance companies that can help customers get answers faster, avoids long phone conversations, and takes the pressure off the contact centres.
Messaging offers numerous benefits:
- The elapsed time to resolution is often faster than a call. While it may take a number of messages to complete an inquiry, the elapsed time to resolution is often faster than calling up, waiting on hold and then finally discussing their issue.
- It’s more convenient for customers. Millions of customers already use messaging apps like WhatsApp daily, so the format is familiar to most customers. Because messaging lets customers interact asynchronously with representatives, they can initiate and respond to inquiries at their own pace, giving them greater control over their own time.
- Contact-centre agents can handle multiple conversations. It’s not just customers who benefit. The costs of servicing customers by telephone are substantial. With phone-based customer service, representatives can only handle one caller at a time. Messaging lets each agent help multiple customers at once, creating efficiencies and (in the longer term) cost savings.
- Customers’ privacy is preserved. A representative may be talking to other people simultaneously, but there’s little risk of exposing sensitive information. Best of all, everything still feels like a one-on-one, personal conversation, putting the customer at ease.
- Messaging avoids routing calls to agents’ homes. Messaging is easier for contact-centre agents to support from home than phone calls, because many agents are having to deal with young children or pets in the background at home.
- Publish detailed pandemic guidance on your site and apps
The first step to helping customers, and reducing pressure on contact centres, is to ensure that customers can access as much up-to-date information about the effect of the pandemic on their finances as possible through digital touchpoints, particularly your site, apps, email and alerts. Specifically, firms should:
- Put Coronavirus guidance front and centre on site and apps. Include detailed information about what the pandemic means for customers on site and app home pages, as banks like Commonwealth Bank of Australia and Royal Bank of Canada have already done.
- Use alerts, email and social media to let customers know. Use outbound messaging such as SMS alerts, push notifications and emails, along with social media posts, to let customers know about the steps you are taking and how to reach the guidance digitally.
- Provide as much information as you can. The eligibility criteria for government support, or mortgage payment relief, are typically complex as governments and financial firms are trying to direct help to those who need it most.
- Update your pandemic guidance daily. Government measures to support workers and businesses are evolving rapidly as the economic crisis unfolds. Chief Marketing Officers should take direct responsibility for ensuring that accurate guidance is being conveyed to customers and that that guidance is updated rapidly as the crisis evolves.
- Make clear which channels are secure, and which are not. Some channels aren’t suitable for exchanging sensitive personal financial information. Make sure it’s clear to customers which channels can be used to exchange private information, and which can’t.
- Don’t tell customers to call. A few CEOs, such as Bank of America’s Brian Moynihan, have encouraged customers to contact the firm by phone, seemingly unawarethat their contact centres are struggling to cope with the current volume of calls. The right instinct, but the wrong message. Instead, encourage customers to use digital touchpoints.
- Train and equip employees to serve customers from home
Firms need rapidly to retrain and equip employees, notably branch staff, so that they can support customers from home:
- Equip employees to work from home. Straightforward steps like issuing employees with laptops and giving them secure, remote access to contact-centre systems will enable them to work from home. That may mean buying 100s of laptops and having them shipped directly to employees’ homes.
- Establish a virtual contact centre. Firms that have not done so already need urgently to set up a virtual contact centre by putting in place secure remote access to the CRM, call-routing, customer identification, knowledge management and other systems that front-line employees need to get the right answers to customers while working from home.
- Retrain branch employees as contact-centre workers. Most banks have both a sudden drop off in branch use and a sharp increase in contact centre volumes. The obvious solution is to train branch employees to support customers remotely. Set up online training courses to teach branch employees how to serve customers using contact-centre systems from home.
- Use live streaming for insurance loss adjusting. Social distancing and lockdowns make home visits by insurance loss adjustors more difficult or impossible. Insurance companies need to adopt approaches like video streaming to help settle some claims, particularly home insurance. Startups like Spain’s Bdeo, the UK’s Tractable, and Librestreamin the US can help.
- Work with partners to rapidly deploy messaging
Secure messaging and chat on digital touchpoints will help to triage incoming customer questions, helping firms support the most complex and urgent problems with human help. There are two easy ways to get started with messaging:
- Work with specialist messaging vendors. Messaging providers like LivePerson, Nivo, Qwil Messenger and Unblu offer mobile Software Development Kits (SDK) that can be quickly integrated into existing banking or insurance apps. This routes to a back-end user interface that messaging agents use to communicate with customers. Because all messages are recorded, it’s both compliant and easy for a different agent to pick up the thread if the conversation becomes asynchronous.
- Use digital platforms like Google and Apple. Google, through its Rich Communication Services (RCS), and Apple, through Business Chat, have both started to offer richer content distribution via messaging. These can either be used standalone from Messages, or via integration with mobile apps. Apple Business Chat also lets firms replace ‘Call Us’ links in adverts on search sites or other media such as Facebook with a ‘Message Us’ option.
- Deploy Chatbots to serve customers
Chatbots offer simple conversational interfaces that let customers use natural language to get answers. Many financial firms like Bank of America, Bradesco Bank in Brazil, Commonwealth Bank of Australia, Lloyds Bank in the UK and US insurer Lemonade are already using chatbots to handle simple customer enquiries. Most chatbots today are rule based, because they are quick, easy and relatively cost effective to deploy.
To get started:
- Consider using specialist vendors such as Finn.ai, Kasisto, Marley, Nuance, Payjo and Personetics, that can help financial firms to get up and running quickly.
- Use frameworks to get started such as Amazon Polly, Dialogflow, Facebook Messenger, and Wit.ai that will speed up implementation.
- Analyse keywords to find relevant answers. Much like how search engines work, rules-based chatbots analyse keywords in text and then take the conversation flow down a predefined route.
- Use visual interfaces to help customers complete tasks. Where there is a visual user interface (UI), such as a website or mobile app, you can use buttons to help customers complete a task in a few clicks. Beware that this is fine where there are a finite number of options for customers but becomes confusing when the number of options is higher, and the complexity is greater.
- Use chatbots to triage for human agents. Rule-based bots can also be great for triaging and collecting information that can be used by an agent to resolve customer issues more quickly.
- Ensure that there’s an escalation protocol. It’s crucial that you put a human escalation protocol in place so that contact-centre agents can take over the conversation, without loss of context, to answer the questions that the chatbot can’t understand or answer.
- Continuously improve your chatbot. Put in place a feedback loop so that you can continually assess the relevance and effectiveness of the responses the chatbot is giving to your customers.
- Introduce AI to make your Chatbots smarter
To move from rule-based chatbots to something that can help answer a wider range of questions, creating more benefits for both customers and the business, you need to create an intent-driven system that uses natural language processing. Various forms of artificial intelligence, specifically natural language processing (NLP), natural language understanding (NLU) and machine learning (ML), can help chatbots understand customer intent and anticipate customers’ needs.
To develop smarter chatbots:
- Use data collected from messaging and call recordings to train your bot. You can train systems to understand a customer’s intent from the many different ways that customers ask the same question, and then to apply the correct treatment. That treatment could be an answer, a link to an answer on your web site or a hand-off to an agent.
- Ask questions to understand customer intent. Triage problems by asking customers a series of questions to help get to the root cause before handing this information off to an agent who can work on the issue before contacting the customer through messaging. This is known as ‘hand-off by design’. It also means for certain intents, such as ‘mortgage-holiday’, you can route customers directly to an agent specifically trained on the subject matter.
- Start with simple conversations. It’s impossible to train systems on everything from day one. Start with high-frequency, simple customer requests. Gradually build the intelligence to handle more complex scenarios to reflect what works and how customers respond.
- Keep analysing conversation data to improve your Chatbot
It’s important to continually analyse what questions customers are asking and how they ask them. This way you can train your system to have a better understanding of the customer’s intent and can then apply the correct treatment. That will enable you to improve how many conversations your chatbot can resolve: its ‘resolution capability’.
To do that:
- Train conversation analysts to teach the chatbot. Set up a team of people whose job it is to analyse the conversations and see which ones work well and which ones seem to confuse the chatbot. These can then be corrected for future conversations, by adding new intents and dialogue into the system.
- Resource these teams from your contact centres. You should be able to free up staff once customers switch to messaging. This is where you will find people who really understand how customers talk. The dialogue flows that your chatbot is trained on should be as natural as possible.
- Continually monitor how effective the chatbot is. Measure ‘containment’, which is the percentage of conversations the chatbot can resolve, as well as ‘able to help’, where the chatbot understands the customer intent and applies the correct treatment. Your goal should be to improve both of these every day.
Benjamin Ensor is the Director of Research at 11:FS. Terry Cordeiro is a Product Manager at 11:FS.