A CRM software provider looked to Frame AI to help them use their customer voice to improve customer sentiment. For many years, the company’s Support team was very proud of industry-leading CSAT scores until their score dropped, and they couldn’t figure out why. “We had a very jarring experience where our Support quality quickly devolved from something the entire organization touted to a liability without an obvious fix,” says the company’s VP, Global Technical Support & Services.
“All of our collective reporting resources and hours of meetings couldn’t tell us why our most important metric wasn’t telling the story we needed to hear. We realized we had to take a hard look at our operations and learn from our raw customer voice, where we could improve.”
Before Frame AI, the Support team followed a first-in, first-out queue logic across the >100k tickets they see each month across email and chat. The company had a basic CRM-Helpdesk integration so that support agents had some basic context about with whom they were interacting. However, CRM data had no formal bearing on the resolution process. Each week saw it’s a fair share of fire drills, with tickets that ended up being considerably more urgent and more complex than anticipated. Even with a detailed tagging schema, the tags failed to sufficiently capture tickets’ substance in a way that the company could operationalize.
With Frame AI, the company quickly learned that they were not prioritizing tickets effectively. Frame AI’s automatic sentiment reporting across 100% of conversations highlighted pervasive gaps in the company’s resolution process. The company learned that “it wasn’t clear to our agents which issues needed to get escalated directly to engineering. When we were escalating issues, we were waiting way too long to do so, and since we lacked a scalable way to provide context for the conversation, customers had to repeat themselves.”
The substance of the ticket now plays a key role in the company’s resolution process. Using a custom, language-driven scoring mechanism, Frame AI passively applies a score to every single ticket. The score helps the company understand every ticket’s relative urgency and complexity so that the next action gets automatically assigned. CRM data also now actively influences where a ticket falls in the queue. For example, a ticket that meets urgency criteria defined by the customer’s language and CRM meta-data triggers a global alert to ensure that an agent answers immediately. Language suggesting that the customer is still stuck or confused can suggest complexity, triggering a follow-up task to sync with Engineering. As a result of the upgraded resolution process, the company has reduced ticket resolution times by 17% and sees half as many risks surfacing across conversations requiring escalation.
“Frame AI has brought unprecedented clarity to our operations. It’s given us the confidence to iterate and experiment, beginning with small tweaks to our processes. With scores available on every single conversation, we know we can measure the impact immediately and change direction if need be.”
The other dynamic at play here is that “our Support team does so much more than troubleshoot password resets. Our product’s value is ensuring that end users can find answers to the questions they have, and Support plays a key role in delivering that. We work so hard to delight our customers, so shoving a survey at them following what may have been a make-or-break moment in our relationship with that customer feels like the wrong approach. Now, we use CSAT less frequently and with better targeting and don’t need to trouble our customers to figure out what’s going well and what isn’t.”
A year into adopting Frame AI, the company has regained its reputation for industry-leading CX, and has also successfully monetized a new Support offering.