A virtual agent is an excellent way to elevate your company’s customer experience. However, if a virtual agent does not have the basic traits needed for success, it can do more harm than good.
Our team performs chatbot UX audits regularly and uses our standard 6 point review process to ensure every bot has the best user experience possible.
A virtual agent needs to be easy to find for your customers. Some criteria for ideal virtual agent placement include:
- Funnel and webpage traffic: Does the virtual agent direct your customers where you want them to go?
- Virtual agent purpose and topic coverage: Is the virtual agent on the social media platform or website page that best aligns with its capabilities?
- Reactive and proactive virtual agent: Does the widget for the virtual agent appear at the right time in the customer journey?
- The right channel(s): Is your virtual agent where your customers are?
- Leverage usability testing to make data-informed decisions: Is the channel or website page the virtual agent is located on aligned with customer needs and data?
A virtual agent needs to have a clear identity and personality that aligns with the brand it represents.
- Perception: Customers will draw perceived personality traits from your virtual agent, whether you design for it or not. Be intentional about your virtual agent’s personality so that it can produce the best results for your company.
- Connection: Customers find it hard to connect with a ‘lifeless’ design, product, or code. The right persona will trigger an emotional connection to the virtual agent and acts as a competitive differentiator.
- Higher adoption rate: A virtual agent’s personality is a key driver of customer engagement and adoption.
- Human behaviour: Customers prefer the virtual agent interface to behave like a human.
A virtual agent needs to fulfill customer needs.
- Resolution time: A virtual agent is only useful to customers if it can efficiently and effectively resolve user issues.
- NLU design: Enhanced NLU design allows a virtual agent to capture user intent and specific details to avoid asking users for information already provided.
- Transactional: Users want to be able to complete actions in the virtual agent environment. Users often find self-serve FAQ’s to be the biggest challenge with conversational agents.
A virtual agent needs to be both intuitive and easy to use.
- Informative greeting: Clarify the virtual agent’s abilities in its greeting to the customer. The greeting should be clear about the virtual agent’s purpose and scope, i.e. problems it can solve, ability to redirect to relevant content or agent, etc.
- Menu options: Menus should be clear and have a flow that includes sub-options when applicable.
- Defining usability: Discerning if users know what to say or do at each interaction turn helps ascertain usability.
- Make data-informed choices: Analytics on abandonment and confusion rates at specific points highlight potential design improvements.
A virtual agent needs to be easy to understand, and trustworthy enough that users trust it with their queries and secure information.
- Incorrect redirection: If a virtual agent often misunderstands customer queries and redirects them to the wrong place, users will not trust it to solve their problems.
- Technical ‘bugs’: If the virtual agent often expresses incomprehension when a customer follows prompts, such as clicking on menu buttons, users will lose trust in it.
- Broken links: A virtual agent should never redirect a customer to a broken link.
- Secure information: A virtual agent that collects secure information from customers should ensure that the user is aware their information is protected.
A virtual agent needs to provide value to the customer using it, and engage them in a personalized manner.
- Value addition: The value of a virtual agent is derived from optimizing customer engagement and NLU performance.
- Leveraging virtual agent “memory”: Storing data points that were previously shared by users, rather than asking for those details again, greatly enhances the value of a virtual agent.
- Understanding customer needs: Training a virtual agent to recognize the true intentions behind the many ways customers express their needs is critical for successful implementation and maintenance.
- Knowing the customers: To meet customer needs, it’s important to steer away from “generic” responses, and offer a more personalized experience by leveraging customer data, past conversations, etc.
Let’s supercharge your virtual agent together, contact us today.