Adrian McDermott, CTO, Zendesk: What CX Teaches Us About AI And The Future Of Work

AI Trends and Enterprise Transformation

• Zendesk serves 100,000 customers, with approximately 15,000 currently utilizing generative AI features, a capability the company has been shipping for nearly a decade.
• Early industry narratives suggested AI would eliminate customer service jobs; however, Klarna CEO Sebastian Siemiatkowski recently nuanced this view, confirming the company is rehiring humans for high-value VIP support after initial automation.
• AI represents a fundamental shift in the "means of production" for customer service rather than merely adding new communication channels, distinguishing it from previous disruptions like telephones or the internet.
• Historical precedents, such as the introduction of ATMs and Amazon's deployment of fulfillment robots, demonstrate Jevons' Paradox where increased efficiency leads to higher demand and expanded human employment rather than job elimination.
• Zendesk data indicates that 54% of their customers contacted support last year, representing 94% of revenue value and 95% of growth, identifying support as a critical engagement hub for top-tier clients.

Strategic Implementation and Ticket Analysis

• Analysis of 33 million Zendesk tickets categorized into 2,400 intents revealed that 71% of inquiries represent internal business failures (e.g., product defects, billing errors) rather than external service needs.
• The remaining ticket volume was split between low-value informational queries (e.g., order status) which automate easily, and a 11% segment of high-value interactions requiring human empathy for churn prevention and upselling.
• AI agents function as a low-code environment for "citizen developers" (customer service agents), allowing them to write procedures and build integrations to fix upstream business problems without waiting for engineering resources.
• Zendesk's internal survey data shows 51% of users prefer resolving issues with bots for speed and reliability, a figure projected to increase annually as psychological preference shifts toward machine resolution.

Pricing Models and Product Roadmap

• Zendesk has transitioned toward outcome-based pricing for AI agents, launching a model that charges $2 per resolution rather than per seat, effectively decoupling revenue from the number of human agents.
• The company is developing spoken support capabilities for Q3, integrating speech-to-speech technologies from ElevenLabs and OpenAI to apply existing text knowledge bases to voice interactions.
• Future pricing experiments include a potential "license every person" model where customers pay a fixed price for human and automated resolutions, removing traditional seat-based revenue structures.
• Zendesk plans to maintain a competitive stance against AI-native startups by democratizing access to technology while adhering to customer comfort levels regarding risk and GDPR compliance.

Outlook and Risks

• The customer service market is projected to shift from a $500 billion human-dominated sector to a $250 billion software-dominated sector by the end of the decade, with IDC estimating the current total market at $600 billion.
• Primary risks for AI adoption include customer hesitation due to skill gaps and risk aversion, requiring vendors to balance innovation with trust-building and education.
• The speaker maintains a macro-optimistic view on employment, acknowledging that while individual transitions involve winners and losers, the net result will be a move toward more fulfilling work utilizing human judgment and empathy.
• Zendesk advises organizations to treat customer support as a source of revenue and engagement, using AI to offload repetitive "fix-it" procedures while redirecting human talent toward high-value relationship building.