Support ticket volume is an underrated UX metric. If your product doesn’t make itself clear, users are going to have questions and submit support tickets. Responding to those tickets takes time and resources, but thoughtful design can answer questions before users even know they have them.

Support needs differ by industry and company maturity level. An ecommerce company at any growth stage has an average ticket volume of 0.1-0.4 tickets per customer/month. Ecommerce companies can also get away with longer response times— up to 72 hours on average. Early stage SaaS companies, however, have more requests and shorter response times. A new SaaS company has an average of 0.3-0.8 tickets per customer/month. For a mature company, that average drops lower to 0.1-0.3 tickets, but fast responses in under a day also become the norm.

Solving support requests costs money. SaaS companies in North America spend an average of $15.56 per ticket when factoring in software costs and employee wages. If your company is receiving 1,000 tickets a week, that adds up quickly.

Braincuber, an AI-driven agency, helped a 7-figure Shopify brand get their weekly support ticket volume down from 1,900 to 763. The brand used Zendesk for ticketing and had 11 employees managing customer support. Their small team faced significant attrition. 3 of the top performers quit within half a year due to burnout from being asked the same questions repeatedly.

The first thing they tried before seeking help was a chatbot. All it could do was answer a small fraction of questions before handing them off to employees with no context. The employees called it “the confusion machine.”

The generic chatbot approach didn’t work. They partnered with Braincuber to build a custom AI agent. Instead of a simple decision tree, the new agent was created using training data from Zendesk, the user’s live order data from Shopify, the returns management app, and the company’s product knowledge base. This resulted in an agent capable of taking user context into account when providing support. It was also able to ask clarifying questions and understand when to escalate a ticket to the human employees.

After 90 days, the new agent helped decrease support ticket volume by 60%. It also reduced the first response time to mere seconds and lowered the cost per resolved ticket by $6.50. Because only some tickets required human intervention, they were able to operate without replacing the agents who had quit due to burnout.

Personalization. The agent replaced a generic chatbot and provided a custom user experience with proper context. Not only does this personalization feel more premium, but it also makes it possible to resolve more complex issues without having to escalate. Good self-service is a boost for UX, as people are busy and do not want to have to wait around for help.

The new agent drastically reduced the number of tickets that went to the human support team by handling the most common requests, like returns and order status inquiries. Over time, this generates significant savings.

Implementing the agent cost $43,000. The brand made its investment back in just over 6 weeks.

If you found this useful, forward it to someone who is interested in what design can do.