9 minute read · Published April 13, 2024

How to improve your ticket deflection AND improve user experience

Latest Update May 10, 2024

Every business wants to provide a great customer experience and grow sustainably as fast as it can.

Unfortunately, as many businesses grow, those two goals can get in the way of each other, and greed takes over. This can be especially true of customer support.

Have you ever called a support line only to find yourself on the line with someone in a call center who has no real agency to fix your issue or little context on the reality of the user's perspective?

Have you ever been in an endless cycle of waiting and failed support over chat?

You're not alone.

74% of Americans say that they’ve experienced bad customer service in the last year!

Solving the support crisis

Companies can try to solve this issue by employing quality support agents who are empowered to help.

But that can be expensive and unrealistic for small startups with low capital and/or massive companies with millions of customers.

That's why there's a huge focus these days on ticket deflection.

You might have heard about Klarna’s AI support agent, which “performs the equivalent job of 700 full time agents” and helped “customer resolve their errands in 2 min vs 11 min” and handled over 66% of Klarna’s total support inquiries.

This AI agent deflected tens of thousands of tickets away from needing to be serviced by human agents AND decreased customer resolution time.

This is all part of a general trend towards AI-powered customer support, which is actually faster and more cost effective than traditional human support.

Let's dig into how you can best deflect tickets while maintaining a quality user experience.

Ticket deflection basics

Traditionally, ticket deflection has come in many forms, from creating healthy and robust self-service options like a knowledge basis or static chatbot, to unhealthy and unethical dark patterns that make it nearly impossible to actually get in touch with a human being. Let’s quickly review these:

Help center

You've got standard things like a help center or knowledge base that has an article repository that folks can tap into and read to answer their questions. Your team creates this content, organizes it, and then links to it within your product to help your users.


Some companies also have user-generated content hubs like a forum or discussion board. Others have users’ who create many videos and images to assist other users. The benefit of this is that your users create valuable content for other users, and as more and more content is produced, the quality usually gets better and more robust.


Over the past 10 years, chatbots have become increasingly popular as a method to increase user support speed and satisfaction while reducing the number of tickets agents have to handle.

These mostly operated using a decision tree, much like when you're calling into phone support and you press different numbers to select what you want to do based on the options.

Chatbots have certainly come a long way, and the original method of a static logic tree set of interactions does not hold much weight in 2024 when AI advancements are all around.

Feedback and improvement

Finally, the best companies make sure not just to deflect tickets and answer questions but to collect lots of feedback both quantitatively from all those interactions and also qualitatively to understand the gaps in their support flows.

All of these efforts can help you deflect tickets!

How to measure ticket deflection

Ticket deflection is a really important metric for measuring the success of your customer service efforts. It's a gauge of your ability as a company and support organization to resolve tickets through your self-service options before they reach the level of a human agent.

Measuring your ticket deflection rate is fairly simple.

Ticket deflection rate formula

One thing that's important to ensure is standardized across your organization is how a deflection or a self-service option is defined. The most common ones will be through your knowledge base or a chatbot, but either way, just make sure it's consistently measured throughout the lifetime of your data.

What’s a good ticket deflection rate for SaaS?

There's no one average for SaaS businesses for a ticket deflection rate, as your vertical and Company size and support structure can make things very different.

A ticket deflection rate of around 20% to 30% is considered solid for most SaaS companies. Companies with a purposeful plan and a highly effective support team could see rates as high as 40% to 60%.

Why does ticket deflection matter?

If you're still not convinced that ticket deflection should be a priority, let's examine some numbers.

Research has shown that resolving a support ticket costs an average of about $15. That includes your team's time to read, analyze, solve, and respond to users, as well as any added costs.

That doesn't seem crazy, but if you consider the volume of tickets that even small organizations receive monthly, it definitely starts to add up.

Let's say you've got 50,000 MAUs on your platform.

2% of them file a support ticket a month.

That’s 1,000 tickets a month!

That’s $15,000 a month in spend!

And so that’s $180,000 a year!

Now, imagine you’re able to increase ticket deflection by 40% with some AI tooling.

That’s $64,000 in savings that boosts your net profit!

Why ticket deflection matters

Finally, imagine these numbers for large SaaS companies! This spending can go into the tens of millions.

Now, before you go and do everything in your power to deflect tickets at all costs, let’s discuss something important.

Ticket deflection is not always a good thing … really!

Let's be clear: ticket deflection is not always a good thing. As we discussed, the reason your tickets are being deflected is very key, and you need to understand this before you can determine whether your strategy is working.

Let's say you set up a new chatbot and saw your ticket deflection rate rise by 40%.

That looks great on paper, but if you dig into it and realize that the chatbot is super shoddy, not directing your users to any available resources or answering any questions, and folks are just leaving the flow, you're setting yourself up for an angry and frustrated user base.

Obviously, that's not the desired end state here.

So, how do you balance your need for better ticket deflection with maintaining or even improving your user experience?

How does AI play a role in ticket deflection? Chatbot deflection rate increased

One of the best ways to bridge this gap is to introduce an AI Copilot onto your SaaS platform.

We’ve been focused on developing a best-in-class AI Copilot that can answer questions quickly, dynamically, and in a persona that resonates with your users

Copilot can be trained on all of your help documentation, knowledge base articles, and more, and then will dynamically respond and interact with your users' questions and queries.

What's really powerful about this is that users can ask questions in their own natural language and quickly get answers in natural language as well.

You can create static flows for repeat and common questions where it always sends them to a specific article, or you can let the AI Copilot run its course and train itself over time.

It's fairly plug-and-play, which is powerful for many folks. You can be up and running in days, not months.

Unlike your support agents, it's available 24/7 and continuously learns from user feedback. After it suggests a specific article or answers a question, you can set it up to query for thumbs-up or thumbs-down feedback, and over time, it will learn.

In addition to that, your product team can review Copilot data in your dashboard and find all of the fallbacks, which is where it wasn't able to find an answer. It fell back on your designated fallback route, like an agent flow or a specific article or a phone number, and then work to improve your responses.

Between this automated learning and this manual improvement of your help documentation, guides, and more, you can supercharge your customer experience.

Why combine Copilot with human agents?

The whole point of your customer service efforts is to improve your user experience and satisfaction.

When support is fast and personalized through Copilot, it has an inherent smoothness that users appreciate and come to expect.

This way, you're not only deflecting tickets, but you're also increasing user satisfaction and happiness, which should be your overall goal.

However, it's not necessary to completely eliminate all of your support agents. They can be a great fallback and useful for more complex cases.

Copilot helps support agents

In fact, you'll not only reduce your cost per ticket and total time spent on tickets but also increase your agents' happiness if their time is spent on more involved and intense tickets, which will produce more dopamine and satisfaction.

Now, that’s not to say that incorporating an AI copilot into your support team won't be a challenge, as it can certainly provoke some anxiety amongst your team members about the potential irrelevancy of their role.

But you can make it work if you have a clear strategy to emphasize the support agents' instrumental role in producing a high-quality experience for the more difficult tickets.

Balancing ticket deflection strategies against your UX

While ticket deflection is an enviable goal and something we should all strive for, it's important to balance it against the overall user experience and happiness. Don't look at it in a vacuum; consider your ticket deflection rate trends against your UX.

One of the best ways to deflect more tickets while also increasing user satisfaction is to use Copilot, where natural language inputs produce quick and fast answers and directions.

Within CommandBar's Copilot, you can even launch nudges, tooltips, or product tours for specific queries, creating a smooth overall app experience.

CS teams need to monitor a lot of metrics, and all of these can feed into the overall performance of both your ticket deflection strategy and your team’s performance as a whole:

  1. Usage Rates: This measures how frequently customers use self-service resources. High usage rates can indicate that customers find these tools valuable and accessible.
  2. Completion Rate: This tracks whether customers are able to resolve their issues through self-service tools without needing to contact support. A high completion rate suggests that the self-service options are effective.
  3. Customer Satisfaction (CSAT): This metric assesses customer satisfaction with the self-service experience. It's usually measured through surveys that ask customers to rate their satisfaction after using self-service tools.
  4. Net Promoter Score (NPS): While NPS is a broader measure of customer loyalty and satisfaction, it can be segmented to assess the impact of self-service experiences specifically.
  5. Self-Service Score: Some organizations calculate a specific metric that evaluates the overall effectiveness of self-service tools by combining various factors such as usage rate, success rate, and impact on customer satisfaction.
  6. Average Resolution Time: This metric measures how long it takes for customers to resolve their issues using self-service tools. Faster resolution times can indicate more efficient and effective self-service options.
  7. Bounce Rate: In the context of online self-service tools like knowledge bases or forums, bounce rate measures how quickly customers leave the site after viewing only one page. A high bounce rate might suggest that the content is not meeting customer needs.
  8. Escalation Rate: This metric tracks how often customers escalate their issues from self-service to direct contact with customer support. A lower escalation rate generally indicates that self-service tools are successfully resolving customer issues.
  9. Traffic Sources: Understanding where your traffic comes from can help you identify which channels are most effective at directing users to your self-service tools.
  10. Feedback and Comments: Qualitative feedback from customers about their self-service experience can provide insights that are not captured by quantitative metrics.

What factors into ticket deflection rate?

  1. Product Complexity: Companies with more complex products might have lower deflection rates because customers are more likely to need personalized assistance. Simpler products, on the other hand, often have higher deflection rates as many issues can be resolved without human intervention.
  2. Quality of Self-Service Tools: The effectiveness of a company's self-service tools directly impacts ticket deflection rates. Better quality and easily navigable self-service options generally increase deflection rates.
  3. Customer Demographics: A tech-savvy customer base is more likely to use self-service options effectively, leading to higher deflection rates. In contrast, less tech-savvy users might prefer direct interactions with customer service teams.
  4. Continuous Improvement: SaaS companies that regularly update and improve their self-service tools based on customer feedback and usage analytics tend to increase their ticket deflection rates over time.
Copy icon
X logo
LinkedIn logo

Zero fluff.

Our stories 🤝 your inbox

Email Icon

Up Next

Increasing user engagement the right way from the ground up

Never miss another
fabulous article.

Weekly digest in your inbox.

Email Icon
Continue Reading

Up Next

The subtle art of the changelog

Read Now

What's hot