Increased customer engagement, time savings, cost reduction, and customer satisfaction are factors driving many organizations to deploy customer service bots.
And it’s easy to see why – customer service bots, such as those built on the Dynamics 365 Virtual Agent for Customer Service framework, enable scalable and cost-effective solutions for businesses to fulfill this drive.
The customers for these bots could be external customers or employees within a company, or any relationship between an end user (customer) and a product or service.
But in order to build a successful bot there are a number of considerations and concerns.
In this blog, I’m going to go over the importance of identifying your business processes – and therefore what needs to be automated – and what you should consider when thinking about who your customers are.
Identify your business processes
Before you can start planning a bot, you need to have an idea of what business processes you want the bot to handle.
The very first step is to look at identifying those business processes that can be fulfilled with the least amount of human interaction, but that provide users with a solution to their queries.
For example, a retail organization can review the various business processes within their customer’s shopping journey; an example is shown in Figure 1.
Figure 1: Example customer journey showing retail business processes
In this example, a customer support bot could help users find the right product at the best price. It could also help in knowing if a product is available in inventory depending on the shipping type.
If you are a subscription-based or service-based business, your business process could be similar to the cycle described in Figure 2.
Figure 2: Subscription-based and service-based cycle
For these types of businesses, a customer service bot can play a part prior to starting the service by helping customers choose the right plan, starting or stopping the service, and usage of the service including helping with service interruption or outages, billing, and payment.
These are just two common examples. Other examples include help desk support for an IT department in an organization, in which case the bot could assist with common troubleshooting steps, or a government agency that needs to provide licenses and permit forms to citizens.
Know your customers
Your customers, their segmentation, and personality traits play an important role in successful usage of your bot. Some of these factors include:
- Age group – Do your customers fit within defined age groups?
- Type of users – How proficient are your customers with the product or service?
- Geographic spread – Where are your customers located?
- Language – Do your customers speak the same language, or do you need to support multiple languages?
Each of these customer segmentation traits impacts the design, length, and types of interaction that need to be automated.
For example, some segment of customers will only use a support bot after they have exhausted all self-service options. On the other hand, some customers who will wait for a solution after a lengthy conversation with a human agent might not exhibit the same level of patience with a bot, and thus it’s important to deliver a solution quickly.
The geographic spread of your customer base can impact the design of the bot, as solutions within the topic could vary within geography. For example, solutions for payment acceptance problems will differ from country to country due to different business rules and payment acquirers across countries.
In addition to who your customers are, it’s important to analyze what your customers care about the most. Similar to the Pareto principle where “80% of problems may be caused by as little as 20% of the product”, our experience with Dynamics 365 Virtual Agent for Customer Service bots has shown that 80% of customer queries are related to the following top 20% of topics or features:
- Most popular programs
- Most marketed or newly launched
- Most problematic business process of your business (for example, visibility of detailed order tracking, real time inventory visibility across multiple channels, payments and billing)
- Complex to understand for end consumers (for example, complex licensing or pricing plans, complex business rules that are not easily understood)
- New releases or feature updates
Every business is different. Hence, it is imperative to figure out the right business processes for your Virtual Agent. Start with building these top 20% of topics that cause 80% of problems to drive maximum containment with best value.
In this blog I’ve gone over some of the considerations when first designing or thinking about launching a customer service bot framework. In particular, it’s important to identify the business processes that you most want to automate, and to understand which customers you need to target based on your product.
Stay tuned for more blogs in the future about further considerations when designing a customer support solution with bots on the Dynamics 365 Virtual Agent for Customer Service framework!
To learn more, visit: