Gaming companies are some of the most advanced users of Zendesk products, because of their diverse and demanding customer base.

Recently, we surveyed our gaming customers to learn more about their views on AI. Are they using AI or machine learning today? If so, what for? What benefits have they experienced? Do they have advice for others contemplating using AI? And if they aren’t yet using AI, what does their AI roadmap look like?

Their responses indicated that the top benefits of AI so far for these gaming companies are reduced agent email and chats. The respondents also shared some best practices, based on their experience so far.

This article includes the following sections:

• Top benefits of AI so far
• Advice from AI early adopters

Top benefits of AI so far

超过一半的流浪儿g customers who responded to our survey said they were already using or beginning to use AI/machine learning to some degree. Several of those are usingAnswer Bot, which uses machine learning to suggest Help Center articles to customers to resolve their issues.

Responses reflected a focus on automation and deflection:

• When asked about their immediate support focus, of those using AI, two-thirds said their immediate customer support focus relates to increased automation and deflection. Responses included: “Optimizing agent processes,” “Increasing speed, accuracy, and convenience [for agents],” “Improving self-service,” “Automation/deflection,” and so on.
• When asked about how they are using AI for deflection, replies included: “General case deflection,” “An easy way for customers to get answers to known issues or questions,” “Knowledge management,” and “Customer modeling and projection.”
• When asked about how they are using AI for automation, replies included: “reduce repetitive tasks,” “automate basic transaction tasks so agents can focus time on more complex cases,” and “provide a better experience for agents.”
• Other responses indicated that agents aren’t the only focus; survey respondents mentioned CX almost as frequently. For instance, responses included: to “[provide] a better experience for customers,” “enhance human touch”, and “give our players a way to self-service many issues”. Also mentioned: “Aid in team development and training” and “Time and cost savings”.

The top benefits experienced so far from using AI were:

• Reduced agent email or chat volume
• Improved prioritization of agent’s work
• Increased CSAT or NPS
• Increased first contact resolution.

Advice from AI early adopters

Gaming customers we surveyed shared the following best practices:

• 投资于你的内容. Until bots are writing their own content, they are only as good as the content and training data that powers them. “Getting your help content right” simultaneously gives your users a better self-service experience and powers Answer Bot.
  Also, leverage content cues driven by machine learning to identify content that needs attention and review suggestions for specific actions to improve content, such as updating article titles, adding labels to improve search, and archiving articles you no longer need.
• Don’t deceive users. Don't be afraid to be explicit when an AI response is being sent. Don't trick your users into thinking they are talking to a human when they are not. Research shows that customers are increasingly comfortable with self-service and automated interactions.
• Audit user feedback.Review any poor CSAT reviews, then follow up with the users who provided them, and make modifications accordingly.
• Reset metric benchmarks. Automating an increasing number of interactions means that only the most complex issues will make it to a human. This means that your team’s handle times, resolution times, one-touch resolution rates, and so on will understandably increase. Managers will need to reset expectations when a human touch is needed.
• Start sooner.The benefits of incorporating AI/machine learning into customer support are worth it; if it’s not in your roadmap, it should be.