How Agatha helped Mixpanel reduce time-to-resolution by 30%

Mixpanel is a business analytics platform founded in 2009, with 300+ employees and 26,000 customers. It currently has 14% of Fortune 500 as customers and was last valued at $865M.

Meet the Support Team

As a B2B software company, Mixpanel's customer support inquiries are complex because their product is very technical and frequently updating. Mixpanel has a Support Engineering team to respond to customer inquiries. The Support Engineering team’s north star metric is the customer satisfaction score - which it measures through a survey that they send to each customer following a support ticket. Mixpanel is proud to have consistently held a score in the mid 90's. Marco Sanchez Junco is a manager of customer support.

The Challenge

Management on the Support Engineering team wanted to reduce customer support resolution time - which is correlated with customer satisfaction scores. Marco also wanted to help his team stay up to date with all the new features and documentation being built at Mixpanel.


Agatha is the answer engine for the enterprise and leverages AI to index past support tickets and knowledge base articles to help support agents answer questions in a fraction of the time. In Marco's words:

“We’re always curious to see what solutions are out there to help us push the envelope. We did some vendor analysis and reached out to several different companies. Everyone’s doing ML! [But] it became really apparent through our in depth analysis that Forethought was the right solution for us. Forethought wouldn’t just surface old tickets, they suggested answers in natural language. This was something we didn’t really see elsewhere. Existing solutions required Mixpanel to [already] have content for something to be correctly surfaced. But we’re a growing org - content is constantly changing. Forethought was the only one that could keep up.”

Immediate Impact

Marco quickly noticed early in the pilot that Agatha was having an impact on his team and could augment the effectiveness of individual support agents by surfacing the right information at the right time. As Marco observed:

"Someone on our team started working on a support ticket, and immediately saw the answer Agatha suggested. She went in and did a bunch of research. Forty minutes later, she came up with the response she wanted to send to the customer. Turns out, our team member’s answer was almost exactly the answer Agatha came up with![Agatha] has really become something the team can rely on and trust.”

Long-term Impact

Not only did Agatha help improve the effectiveness of the support engineering team, it also became a central part of the everyday life of the team. Marco shares:

“Just from talking to people out on the floor, you can tell that this is something they’ve really come to trust. The team got used to it very naturally and is now baked into the team culture. Team members would joke about Agatha, and reference Agatha as if it was another support engineer on the team."


Just as importantly, the approach the Forethought team took to achieve these results also impressed Mixpanel. In Marco’s observation:

“We saw that [the Forethought team] took user feedback very seriously and were folding it back into the product in a rapid iteration cycle...I’ve gone through this vendor selection process several times now, and I’m very surprised by the receptiveness and responsiveness of the Forethought team. Based on the data, we know this product is having a positive impact.”

Get in Touch with Us