Insurance Chatbot Guide

Chatbots have a strong track record for cutting inbound volumes dramatically. Pre-COVID, they were already gaining traction in the insurance industry: just look at the emergence of AI-driven providers like Next Insurance and Lemonade and insurtech vendors like Trōv and LeO. In fact, according to Gartner: Organizations report a reduction of up to 70 percent in call, chat and/or email inquiries after implementing a VCA. So they’re proven to handle normal volumes. But can they scale up to pandemic demands?


1. Customer Service in the Time of COVID

See if this sounds familiar. At 9am you’re supposed to lead the department stand up and set up your daughter’s Zoom meeting with her teacher. You log her in early and rush to your computer. The first five minutes of the call are shot with audio issues (what else?). As soon as you get the meeting rolling, your two-year-old sends his breakfast flying across the room. All hell breaks loose.

Every day it feels like you hit a new record for the world’s longest queue. You’re busier than the last two open enrollment seasons combined and working at 50% usual capacity. Escalating issues used to be as easy as hailing a supervisor from across the room. Now it’s impossible to resolve anything quickly, with employees getting pulled away during peak traffic hours to deal with their own kids.

Page 3 3 (2)

Just when you think the “new normal” of customer service can’t get any worse, you’re trending on Twitter.

For providers just beginning digital transformation journeys, the pandemic created a perfect storm. A skeleton crew of customer service staff struggling to maintain productivity. Inquiries pouring in at an unprecedented volume. Am I covered for a COVID test? How do I file a workman’s comp claim? How should I adjust my policy if I lose my job?

You may have rather desperately researched ways to offload volume—perhaps you looked into solutions like conversational AI that can automate customer-support interactions. What you’ve found is that many of these automation solutions take the better part of a year to deploy. You need to see ROI like, yesterday. What’s a busy remote support center to do?

Enter chatbots.

Page 3 (1)

2. How Automation Can Help NOW

Chatbots have a strong track record for cutting inbound volumes dramatically.1 Pre-COVID,  they were already gaining traction in the insurance industry: just look at the emergence of AI-driven providers like Next Insurance and Lemonade and Insurtech vendors like Trōv and LeO.

Organizations report a reduction of up to 70 percent in call, chat, and/or email inquiries after implementing a Virtual Assistant solution.1 So they’re proven to handle normal volumes. But can they scale up to pandemic demands?

Right now, the Oklahoma Employment Security Commission is using chatbots to field 60,000+ daily calls regarding unemployment claims.2

If lean government organizations can implement AI systems for its call centers, your company can too.

Insurance Chatbots vs. E-commerce Chatbots

If you’ve only encountered e-commerce chatbots built to accomplish simple tasks—like answering general FAQs and recommending the “perfect” dress— you might be surprised to find the solution so robust.

In general, when we refer to “insurance chatbots,” we’re describing a complete virtual agent that runs on intelligent process automation and conversational AI.

New Project (4) copy

Intelligent process automation combines a set of AI technologies intended to complete human-oriented work at the speed of a computer. Namely, IPA incorporates Robotic Process Automation (robots programmed to knock out manual, labor-intensive tasks) and machine learning (AI that analyzes data and learns continuously to make increasingly intelligent decisions).

IPA is how chatbots automate inbound and outbound interactions, execute dynamic workflows and integrate many different types of tasks across channels and systems.

Conversational AI describes any AI-powered technology that can engage in human-like dialogue. This technology enables chatbots to operate as sophisticated conversational interfaces that can answer or ask questions, complete workflows using customer input, automatically create tickets and update statuses and more.

Whew. That’s a mouthful. But basically, these solutions are smart!

There are a few core functionality differences between automated virtual agents designed for complex insurance workflows and basic chatbots designed to recognize buyer intent.

 Insurance Chatbots  


  • Workflow driven: designed to help customers accomplish tasks, self-service
  • Automates both outbound and inbound conversations
  • Integrates with back-end systems like CRMs and databases
  • Agent provisioning and seamless hand-off
  • Omni-channel operation
  • Customize complex workflows to align with customer journeys
  • Military-grade security and compliance
  •  Fully branded white label experience
  •  Advanced AI/ML capabilities

 E-commerce Chatbots  


  • Conversation driven: designed to provide basic information
  • Operates independently, no agent hand-off
  • Limited integrations (API)
  • Runs on websites or Facebook pages
  • Not white labeled








Why Chatbots are Perfect in a Pandemic

It’s not surprising insurance chatbots gained so much traction in non-pandemic times. Customers are already familiar with them, so the user experience is easy. They sit on the front lines of your website: the main channel your customers use to locate info and request services. Automated bots that integrate with your backend data sources facilitate seamless and speedy customer interactions without any human assistance.

But there are some chatbot functionalities that are particularly well-suited for customer service in the COVID-era:

  • Some bots come pre-trained on insurance language data sets—simply introduce your own branded terms and deploy even faster.
  •  They work 24/7 to offload your call center, supporting customers in different time zones.
  •  They can scale to 10x or even 100x call volume without breaking a sweat.
  •  They provide immediate answers to customers—no one wants to wait on hold for 30 minutes.
  •  White-labeled virtual agents communicate in your brand voice, maintaining the customer experience.

But perhaps most helpful is their ability to deflect incoming calls, both proactively and in real time.

Real-time Call Deflection

What usually motivates a customer to call support? A problem that demands immediate attention. Ironically, solving this urgent issue often requires them to spend an hour on hold.

Now let’s say you’re using automated conversational interfaces to complement your call center. When a customer calls in, your IVR system announces the wait time and gives the option to finish their request via a self-service channel (“if you’d like to start this claim on your mobile device, press two..”). From there, the customer accomplishes their task via a branded experience on their phone without any help from an agent.

The quick help satisfies customers and the lighter workload relieves busy agents.

Proactive Call Deflection

We don’t often take advantage of virtual agents as outbound tools. But in situations where you can predict high call volumes, like a pandemic or natural disaster, chatbots can be used to prevent calls before they happen— either by reaching out to the customer and asking for information or by answering anticipated questions.

Using conversational interfaces to automate proactive outbound comms not only saves agent time, it typically shortens the claims process by at least a few weeks.

New Project (5)

3. Prioritizing Use Cases

When your call center is slammed, taking the time to figure out what exactly is making you so busy might seem counterintuitive. But if you don’t properly diagnose the problem, you can only provide a stopgap.

So, if you’re considering implementing an automation solution and aren’t sure where to start, we recommend two methods for identifying the highest priority use cases.

Page 8 2 (1)

High Volume Drivers

Often high volume drivers are simple queries that inundate your support staff.

Some COVID-related high volume call drivers might include frequently asked general customer service and account management questions. For example:

  • “Are your in-person offices still open?”
  • “Will I get charged for cancelling an appointment?”
  • “Are you lowering your rates since your services are impacted by coronavirus?”

Conversational interfaces can power through these responses in seconds. When you automate high volume call drivers, you eliminate interactions that drain time, budget and employee energy.

High Time Drivers

Next, consider the “sneaky” problems that may not drive a lot of volume but suck a lot of time.

  • Claims filing, adjusting, tracking
  • Policy modifications, consultations
  • Refunds and billing

Communication, already asynchronous by nature, becomes even more protracted between a customer and a brand. You call the customer, the customer doesn’t have all the information they need on hand, the customer calls back, you’re closed. Five months fly by before you finally resolve the issue.

Chatbots compress that asynchronicity by automating outbound and inbound conversations.

Special Resources

Perhaps you considered creating a mobile app or subpage dedicated to COVID-19 resources. A special portal might be a good idea from a customer service perspective, but launching new digital properties requires huge time and resource investments.

Newsflash: customers want to use your services on mobile, yes, but no one wants to install another app. You can deploy low/no-code conversational interfaces to quickly stand up an app-like experience accessible through the browser, and use SMS messaging to reach more customers.

4. Implementation

Automation providers offer varying levels of implementation services, so be sure to assess how much support you’ll need before you sign any deals. Some leave bot set up and training entirely in your hands. Others charge onboarding fees (starting in the tens of thousands of dollars) that depend on the use case and complexity of system integration.

In the implementation steps below, we’re assuming a high level of vendor involvement in order to include action items for both chatbot vendors and insurance providers

Post-implementation steps

  1. [Provider + Vendor] Decide what the customer journey will look like for the target use case. What steps does the user need to take to achieve the task?
  2. [Vendor] Train chatbots to respond to target use cases with contextual machine learning models.
  3. [Provider] Educate agents on how best to work with chatbots to maximize resources.
  4. [Provider] Make a plan for escalating scenarios that can only be resolved by humans. How will agents intervene quickly?
  5. [Vendor] Integrate chatbots with CRMs and back-end business systems for seamless customer interactions.
New Project (4)

Following a few extra steps post-implementation will help ensure success.

Working with Automation Effectively

Need inspiration for combining your human and virtual agents? You’ve come to the right place. Here are a few of the many ways insurance providers can use chatbots to complement their agents.

  • X-ray vision: Modern automation systems provide an “agent view” of the customer’s actions across all channels in real time, with user-submitted data stored instantly in a dashboard. That means the agent doesn’t have to call the customer for information they’ve previously volunteered over email, or wait around on the phone while they search for a policy form from five years ago. The customer can interact with the conversational agent 24/7 on their own time, and with instant 360° visibility into customer workflows, reps can quickly sign off on processes when needed.
  • Co-pilot: In this model, the agent “pulls the trigger” while the chatbot does the legwork of capturing information (acting as the agent’s co-pilot). Imagine that your rep is on the phone with a customer signing them up for a policy. The rep needs a signed contract, so they fire off the document via chatbot while still on the phone with the customer. The customer reviews and signs the doc, all within the conversational interface. The rep can see the signature in their “agent view” and move the process forward.
  • Mediator: Some AI chatbots have keen sentiment analysis that auto-triggers agent intervention when it identifies customer frustration. Intervention is seamless (no “you’re now chatting with a live agent” message to throw the customer off). The chatbot answers the basic questions, preserving the agent’s time, while hailing trained staff to the rescue for interactions that require a human touch.

Page 11 (1)

5. Evaluating Solutions

As we touched on in Chapter 2, not every chatbot is created equal. There’s a world of difference between a solution designed to capture leads and answer basic questions and an enterprise-grade AI solution that can knock out complicated insurance workflows—making your agents’ and customers’ lives dramatically easier. Similarly, as mentioned in the previous chapter, there’s a wide range between a “set it and forget it” vendor who will simply sell you a bot and a hands-on partner committed to improving their services and continuously adding value.

On top of the daunting process of finding the right solution and partner, automation calls for a different approach to customer service management and metrics than you might be used to.

To help you evaluate options, we’re sharing our most commonly asked questions as a conversational AI provider. Use them as a starting point to shortlist the right partner for you

1. How can I tell if chatbots are the right way to meet my goals?

A dependable automation provider will help you establish benchmarks around the current problem and your goals for fixing it. What are you doing to solve the problem today? What’s it costing you (both in direct and indirect costs)? Let’s say you’re overwhelmed with call volume. How many inbound calls do you receive on an average day? An above average day? How much back and forth is happening in these calls?

Then, a good provider will help you set KPIs that assess chatbot efficacy. These metrics might evaluate how many calls are deflected proactively, how many calls are deflected in real time (e.g. encouraging callers to complete their query via a virtual agent) and how many chat conversations are started and finished. Providers should also set long-term metrics around customer satisfaction, such as improved NPS and C-SAT scores

2. How quickly will I see value?

Time-to-value can be 0-3 months after deployment for low-code systems and anywhere between 3-12 months for complex systems demanding heavy IT or professional services.

If you ask about nothing else, make sure you evaluate whether the solution requires a large upfront time and financial investment before your customer service department can feel any relief.

COVID-19 put call centers everywhere in a bind. Naturally you want to get a system up and running quickly. Still, if you don’t weigh the short term business impact against time and financial requirements, you may end up an extra $40K in the hole for onboarding alone.

Determine who will be responsible for implementation. Does the vendor offer set-up consultation, and if so, is it incurred as an additional cost or is it included within the SOW? If the implementation is on you, does the system need lots of IT attention to get up and running?

Finally, keep in mind that time-to-value also depends on the automation platform’s IT complexity. Agents can reach productivity much quicker with user-friendly, low-code solutions.


3. How do you protect customer data?

Data breaches are all too common, and insurance providers are an especially lucrative target for hackers. When you’re hosting millions of data points of sensitive customer information, you need to ensure every partner with access to your customer data follows industry compliance standards (such as SOC 2, HIPAA and PCI DSS).

Imagine that you grant a chatbot vendor access to backend data sources so they can integrate with your CRM. Now imagine that the vendor’s own systems are not set up securely. You’ve just exposed millions of customer records to a potential data breach.

Page 14 2 (1) copy

 Compliance may not seem like an urgent issue right now. Again, while COVID-19 has forced insurance companies to prioritize customer service agility and responsiveness, you still need to confirm your automation partner is committed to your long-term success.

6. Chatbots as Business Continuity Tools

The best time to automate your customer service department was 20 years ago. The second best time is now.

All the insurance companies thriving in the pandemic have one thing in common: they’re 5-10 years deep into their digital transformation journeys. And the companies hit the hardest? Digital transformation deniers or laggards. Right now, the early adopters are grabbing market share from that bottom segment. When a customer can’t reach you with an urgent problem, they’re likely to consider switching to a more accessible provider.

Page 15 2 (1) copy

Not convinced? A 2018 PwC study found that negative customer service interactions can turn away even your most loyal customers.

If you happen to fall in the middle or bottom of the market, it’s not too late, and you’re not alone: loss avoidance is a strong motivator for digital transformation.

Think of automation solutions as cogs in the larger strategy to strengthen your department’s resilience. The more you digitize time and cost-intensive interactions, the more you’ve made yourself flexible and adaptable to business climate changes in general— pandemic or no pandemic.

Page 15 (1) (1) copy

Expanding Use Cases

When you introduce automation solutions like chatbots, you’re creating systems that prevent customer churn against potential disasters, environmental and business alike. You can expand your use of automation and increase your resiliency in many ways, but here are three tactics we recommend.

  • Don’t call me, I’ll call you: Be proactive in deflecting inbound queries. If the customer has to pick up the phone, you’re too late. Customer just had a baby? Reach out and tell them how to add a new family member to their policy (and include your congratulations).
  • Low-hanging fruit: Knock out the quickest automation wins. When you solve one, move to the next-adjacent opportunity.
  •  The Big Three: Automate the top three inbound drivers across all channels. Billing questions flooding your email inbox, live chat and support line? Tackle all things billing, then look at the second highest volume query

The COVID-19 pandemic will subside eventually, but the way we work has changed forever. It’s due time for insurers to think broadly about how they can support their customers whether the call center is open or closed.

7. How to Get Automated in Quickly

Ushur is the complete solution for intelligent automation. Designed for high-contact industries like insurance, Ushur engages customers over AI-powered chatbots, email, SMS and more, using conversational AI and intuitive workflows to understand what people are saying and what to do next.

Ideal for busy insurance companies in the COVID-era, our end-to-end automation platform accelerates time-to-value from months to weeks with features like:

  • Free white glove implementation. Unlike other conversational AI vendors, who hand you a chatbot and expect you to train it yourself—or charge an additional $40K for onboarding—Ushur’s implementation is all-inclusive and led by our engineers.
  •  Pre-trained virtual agents. Chatbots typically need to train on 1,000 machine learning examples before they achieve human-acceptable accuracy. We pre-train our conversational interfaces on a database of common insurance language so they start with 80% accuracy out of the box.
  • User-friendly platform. Our interface uses drag-and-drop tools your team can get the hang of quickly.
  • Customizable workflows. Unique to Ushur is our design-your-own automation workflows. Tailor the customer journey to fit any use case for COVID and beyond.
  • Military-grade security. SOC 2 Type 2 and HIPAA compliant, complete with AES 256 encryption and multi-factor authentication.

Set your customer service team up for success in any environment. Contact us for a demo today.

To learn more about insurance chatbots, read our case study on how a provider used Ushur’s virtual agents to cut short-term disability claims processing from three weeks to one hour.

Looking to automate your customer engagements with a vendor that prioritizes your data security?


About Ushur

Ushur delivers the world’s first AI-powered Customer Experience Automation™ platform that has been purpose-built, from the ground up, to intelligently automate entire customer journeys, end to end. Designed to deliver delightful, hyper-personalized customer experiences through rapid issue resolution and unified, omnichannel engagement, Ushur is the first-of-its-kind system of intelligence. It combines Conversational Automation and Knowledge Work Automation in a No-Code, Cloud-native, SaaS platform to digitally transform every step of the complete enterprise customer experience – from Micro-engagements™ to entire customer journeys.

Backed by leading investors including Third Point Ventures, 8VC, Pentland Ventures, Aflac Ventures and Iron Pillar, Ushur’s Customer Experience Automation™ solutions are currently in production at some of the leading insurance providers across the globe including Irish Life, Unum, Aetna, Cigna and Tower Insurance.

Related Posts

Sorry, we couldn't find any posts. Please try a different search.