OpenNashPodcasts

Topic

Zendesk

18 facts from 1 episode, in date order. These are what speakers said on the shows, each linked to the moment it was said. They aren't independently verified.

2022

In December 2022 Zendesk had about ten AI data scientists and engineers, and Tom Eggemeyer asked them for a roadmap in Lisbon, Portugal.

“I remember this vividly in December 2022 with some of our, uh, lead, um, we had a small group of AI, uh, data scientists and engineers, about 10 of them”

Tom Eggemeyer · Zendesk CEO on AI for Customer Experience | CXOTalk #866, 0:06:23

Undated

Zendesk says its AI Agents Advanced product can resolve more than 30% of customer interactions almost instantly after three clicks and company knowledge are added.

“we can solve almost instantaneously more than 30% of customers interactions.”

Tom Eggemeyer · Zendesk CEO on AI for Customer Experience | CXOTalk #866, 0:01:29

Undated

Tom Eggemeyer says some Zendesk customers reach 85-90% AI resolution rates while customers with complex use cases reach only 20-30%.

“There are some of our customers that are getting 85 or 90% resolutions, but there's some customers with more complex use cases that are getting 20 or 30%.”

Tom Eggemeyer · Zendesk CEO on AI for Customer Experience | CXOTalk #866, 0:02:45

Undated

An Australian e-commerce customer of Zendesk had five different internal policies defining what counted as a return, which hurt AI accuracy.

“they had 5 different policies, for instance, on from different parts of the organization on what constituted a return.”

Tom Eggemeyer · Zendesk CEO on AI for Customer Experience | CXOTalk #866, 0:04:18

Undated

Zendesk launched generative search, which returns a direct answer instead of a list of links, as a free part of its knowledge base product.

“we've launched something called generative search, and this is free part of our product for knowledge bases”

Tom Eggemeyer · Zendesk CEO on AI for Customer Experience | CXOTalk #866, 0:12:03

Undated

Zendesk's quality assurance product scores every human and AI agent interaction for accuracy, tone and resolution using an algorithm other than a large language model.

“we actually score every single AI agent or human agent interaction for accuracy, for tone, for resolution.”

Tom Eggemeyer · Zendesk CEO on AI for Customer Experience | CXOTalk #866, 0:19:23

Undated

Zendesk introduced an automated resolutions pricing model under which it is paid only when its AI resolves a customer's problem.

“we've come out with our new pricing model called automated resolutions. We only get paid if we solve a customer's problem with you, if we resolve it.”

Tom Eggemeyer · Zendesk CEO on AI for Customer Experience | CXOTalk #866, 0:32:33

Undated

Tom Eggemeyer describes customer support software pricing as moving from per-seat to per-interaction and now toward outcome-based pricing.

“In the past, it was seats. We still have some seats. Some companies made the twit- uh, the switch to interactions.”

Tom Eggemeyer · Zendesk CEO on AI for Customer Experience | CXOTalk #866, 0:33:53

Undated

Tom Eggemeyer says some of Zendesk's best customers avoid being named as references so competitors will not learn they resolve 70-80% of interactions with AI.

“some of our best customers don't want their name out and don't want the results out”

Tom Eggemeyer · Zendesk CEO on AI for Customer Experience | CXOTalk #866, 0:31:26

Episodes