Kajabi turns to AI and self-service to meet quadrupled service demand
Kajabi empowers knowledge creators and entrepreneurs to achieve success online. When its business experienced sudden, unprecedented growth, the customer support team saw its tech support inquiries quadruple. To meet the challenge, the Kajabi team turned to AI, machine learning and self-service to help handle the demand. The result: a 100 percent increase in self-service usage while maintaining all-important customer satisfaction scores.
"When you're handling thousands upon thousands of tickets, AI gives you a better way to understand and keep a pulse on what is happening with your customers.”
VP of Customer Experience at Kajabi
“We look at deflection as a way to provide answers really quickly. But we always use CSAT as a counterbalance and monitor that very closely as an organization.”
VP of Customer Experience at Kajabi
Number of students
Number of agents
Increase in self-service
Self-service deflection rate
Everyone is familiar with websites that allow you to buy and sell goods and services. But what about a software platform that can help you sell your knowledge? Enter Kajabi, the leading platform for knowledge entrepreneurs and creators. Experts create online video courses, podcasts, and communities in their specialties, while buyers can learn and adopt new skills and participate in conversations with like-minded learners. From health coaching to dating skills to business growth to leadership training, Kajabi makes it possible to turn knowledge sharing into something sustainable and revenue-generating, for both first-time creators and seasoned entrepreneurs.
Since launching in 2010, Kajabi has empowered over 50,000 knowledge entrepreneurs in 138 countries to serve 60 million students and deliver more than $3 billion in sales. When the 2020 global pandemic upended people’s livelihoods and routines, Kajabi found itself with an enormous increase in members looking to leverage its online platform, nearly quadrupling its customer support queries.
Kajabi’s VP of Customer Experience, Jared Loman, says his team had to react and adapt quickly to the influx of new customers. “It was neither financially or logistically possible to throw more humans at our fast growth. It takes several months before new agents are fully trained, onboarded and ready to make a meaningful impact. We did not have the luxury of time so we had to look at another solution.”
The company had already started looking into AI, machine learning and self-service as part of its forward-looking CX strategy, but the surge in volume pushed the team to swiftly take action sooner than planned. Kajabi explored various customer support AI products and, after much research, chose Zendesk partner Forethought to integrate with the company’s Zendesk CX platform.
Making AI smarter so it can work harder
“To meet the demand, we started implementing self-service and offering one-to-many training webinars for our new customers,” explains Loman. “We used that as our initial mechanism to get customers in the door, get their questions answered as quickly as possible, and hopefully reduce the amount of wait.” Then, they added those recordings to the help center.
Loman says that his team had a bit of a head start in confronting the spike in customer tickets due to the research they had done previously as part of their self-service strategy. Their one-to-many approach already included a way to turn repeat customer questions into answers for everyone. The driving force behind the Forethought AI solution was to leverage AI to make the process of answering customer queries seamless and intuitive for both end users and the company’s support team.
Soon after implementing Forethought, Kajabi began to see results. With Forethought Triage, Kajabi began tagging support tickets by Category and SubCategory fields. This decreased the amount of manual work needed from agents. Forethought Solve, an AI-powered virtual agent, began deflecting customer inquiries directly to existing knowledge base articles so customers could answer their questions without the need for Kajabi support staff assistance.
Another major benefit of the Forethought AI solution has been Kajabi’s increased ability to better understand its customer base, according to Rick Sanchez, Kajabi’s customer support division leader.
“Our partnership with Forethought and the data it provides us with has really enabled us to hone in on what our customers are asking for,” explains Sanchez. “Their platform gives us the ability to take apart every single ticket, look for trends, and find new ways to help our customers succeed. This insight into what our customers are asking for is a huge benefit.”
Video captioning revs the AI engine
A key learning during this time was that the help videos needed to be captioned so that the content could be ingested by the AI machine learning technology. Due to the large number of self-service videos, the need for captioning required a fairly significant investment both in human and financial resources.
Although Kajabi was already video-centric in its help center, the company ramped up the development of self-service informational videos to an even greater degree. At the time, the company had one individual on the customer service team dedicated to working full-time on creating and updating information for the help center. Today, a team of six manages help center content to ensure a quality self-service experience.
More and better content in Kajabi’s help center had a direct impact on customer experience. ”Machine learning and AI is really all about self-service,” says Loman. “The AI component is a search engine, it’s about surfacing the right resources for the user to answer whatever question they’re asking.”
“This was a large task. We have a really big platform containing a website builder, marketing automation tools, customer relationship management, and the educational course building platform,” explains Loman. “So, it’s technically four pieces of software in one. We’re not just documenting one thing, we’re documenting almost four full-fledged feature applications.”
At the end of the day, Loman says the investment in video development and captioning was the right call to reap the benefits of the team’s AI machine learning strategy.
Surprises from AI…in a good way
As Kajabi has grown, the AI strategy has delivered numerous benefits to the company’s customer service team.
“When you have a smaller team, you can easily share among each other the feedback and learnings you are seeing from support tickets,” offers Loman. “But when you’re handling thousands upon thousands of tickets, you’re just not having those same conversations anymore. AI gives you a better way to understand and keep a pulse on what is happening.”
Loman shared that the new AI data has been compelling in a couple of ways. First was affirming some long-held hunches: ”That was really huge for us to gain affirmation and validation that things we thought were the case were actually reflected in the data.” But the data also brought in new opportunities to improve. For example, the Kajabi team was surprised to find that a large number of tickets were coming in for a settings page in Kajabi’s software application. “The amount of tickets that were coming in for that particular area blew my mind because it’s fairly simple and it doesn’t seem like a high traffic area of the app,” says Loman. “So AI exposed an opportunity for us to get sniper targeted on a support area for learning and help that I don’t think I ever would have anticipated.”
Confidence is key for the success of AI and self-service. Kajabi also offers its members the ability to interact with a bot to recommend content. Everyone gets to help make the bot smarter, and the bot returns the favor by only responding with self-service recommendations when it is around 95 percent confident that it has the right answer based upon all the previous iterations.
“Nobody really wants to contact support. They just want to be able to use the application and do the thing that they came in to do,” says Loman. “So if we can take the edge off of that process and make it even quicker, I think we’ve been successful in the self-service arena.”
When Kajabi began offering self-service options, the targeted deflection rate was 9 to 10 percent. With additional optimizations to the process, the rate is now up to 17 percent. “That’s a huge success for us considering the complexity of our software,” offers Loman. “We’ve basically doubled the deflection rate, given our baseline target.”
The team is striving to reach 30 percent, but Loman says customer satisfaction is way more important than a high deflection rate.
“Our customers are number one to us and we have to ensure that they are happy 100% of the time,” says Loman. “We look at deflection as a way to provide answers really quickly. But we always use CSAT as a counterbalance and monitor that very closely as an organization and in our OKR reviews.”
Personalized experiences on roadmap ahead
Loman says Kajabi is just scratching the surface of what is possible with AI and machine learning.
“I’m excited about creating very personalized experiences for our customers,” says Loman. “Machine learning and AI can assist with things that agents cannot do on their own. It can surface information that would take an agent several minutes or even hours to accomplish.”
So not only does personalization promise even greater efficiencies for the company’s customer service agents, but it also demonstrates Kajabi’s continuing commitment to provide the best possible service to its customers.