The UI of Chatbots: Part 2 Balancing assistance approaches

• User-controlled interaction

Users can engage with the chatbot only when they feel they need help, which avoids unnecessary interruptions.

This feels more natural to users who prefer to explore and troubleshoot independently before seeking help.

• Frustration before help

Users might have already reached a point of frustration before engaging with the chatbot, potentially leading to negative emotions and reduced satisfaction with the experience.

If users don’t know that help is available or how to ask for it, they might give up before finding assistance.

• Targeted and relevant assistance

As the chatbot is invoked after a problem arises, its responses are more focused and context-specific, directly addressing the user’s needs at that moment.

This prevents the chatbot from giving information that the user might already know or find irrelevant.

 

• Slower resolution

As the bot only reacts after a problem has occurred, the time between encountering an issue and getting it resolved might be longer, resulting in slower task completion.

• Less intrusive

By being reactive, these chatbots do not interrupt the user’s experience. This can be beneficial for users who dislike pop-ups or unsolicited assistance, especially during tasks that require concentration or personal preference.

• Requires user initiative

Users must take the initiative to ask for help, which can be a barrier for those who are less tech-savvy or unsure how to articulate their problem.

Reactive chatbots are less effective if the user doesn’t know what question to ask or that the chatbot is available.

• Limited context

The chatbot might have limited context about the user’s journey up to the point of interaction, requiring additional back-and-forth communication to understand the issue fully.

• Anticipates problems early

By guiding users step-by-step or detecting areas where confusion typically arises, proactive chatbots can help users avoid errors or inefficiencies, improving the overall experience.

Early interventions can reduce the likelihood of users abandoning tasks out of frustration.

• Can be perceived as intrusive

If the chatbot intervenes too frequently or offers suggestions that are not relevant to the user, it can feel intrusive or annoying, leading to frustration rather than helpfulness.

Users might feel patronised if the chatbot intervenes too early or too often, especially if they prefer exploring or problem-solving on their own.

• Faster task completion:

By providing assistance pre-emptively, these chatbots reduce the time users spend searching for help or making mistakes, leading to faster task completion and a more streamlined experience.

• Risk of overcomplicating the experience

Proactive chatbots might overwhelm users with too much information or assistance when it’s not needed, potentially complicating simple tasks.

Overzealous proactive bots can make an interface feel cluttered or interruptive, reducing focus on the primary task.

• Higher engagement and user confidence

Proactive chatbots can create a sense of support throughout the user journey, fostering confidence that help is available if needed. This can enhance user engagement, especially for complex tasks or unfamiliar environments.

• Context sensitivity

It’s challenging for proactive chatbots to accurately gauge when a user might need help. If they get it wrong, they risk interrupting users who were doing fine, or worse, not offering help when it’s genuinely needed.

• Reduced cognitive load

By offering contextual hints or tips at the right time, proactive chatbots can simplify decision-making for users, reducing the cognitive load and making the overall experience feel less overwhelming.

• ·Might limit user exploration

By offering too much guidance, proactive chatbots might inadvertently limit the user’s ability to explore or experiment with the system on their own, which can be an important aspect of learning or mastering a new tool.