Zendesk's outcome-based pricing balances AI with human touch

AI gives businesses unprecedented opportunities to enhance customer service and experience in ways that were scarcely imaginable just a couple of years ago. The rise of large language models that can interpret requests and vast improvements in natural language processing have enabled new channels and autonomous workforces that can support customers in new and innovative ways.

Over almost three decades, Zendesk has found ways to enhance customer support. That means engaging in continuous learning, refining and expanding knowledge.

This has led to the Resolution Platform. Kellie Hackney, the company's VP for ANZ explained, "It's our purpose-built AI-powered system. Resolution Platform shifts the customer and employee service away from basic ticket management and deflection to guaranteed end-to-end customer resolution."

Hackney said it brings together autonomous agentic AI, back-end system connectivity, and human expertise across those channels. The platform is powered by a continuous learning engine that improves on itself to remove friction and create speed to resolution with the ultimate goal of strengthening brand loyalty for the customer and business.

"For too long companies have been using tools that deflect the customer," said Hackney. "Queries should be resolved in the most effective and efficient way, and we've built the reality that solves for the customer. We want to banish chatbot rage. That's what drives the vision for our autonomous service workforce."

The problem with many bots is that they are rigid and scripted decision trees. Once the customer interaction deviates from the decision tree, the bot process breaks. It doesn't have the memory and the context rich component so it's unable to solve the problem.

"Most people don't hate automation. They hate friction. They hate it when something is hard. They hate it when something is slow. Our goal is to replace that old bot with an autonomous service that has specialised agents that coordinate and act. The back-end system connects APIs, MCPs and agent-to-agent connectivity that unlocks the organisation's back end to give the AI agent the power to solve the problem for the customer," Hackney said.

While the customer service use-cases are easy to identify, Zendesk's platform has been used in some less obvious contexts. For example, Carsales uses Zendesk's Agent Builder to connect backend systems to create specialised billing, claims and collections agents that enable teams to work from a single, governed platform.

Sports monitoring and management platform Catapult, which is used by leading sports clubs around the world including many AFL, NRL and F1 teams, use Zendesk to collect data from multiple systems to give management teams insights into athlete performance in real time.

Carsales achieved a 90% handled rate by cross‑skilling staff and using Zendesk copilot to surface the right knowledge base instantly.

The use of intelligent agents to has driven Zendesk to pioneer a new pricing model. While traditional per-seat licensing remains available, the company has introduced Outcome Based Pricing.

"Our core belief is that customer loyalty is retained when the query is resolved," said Hackney. "So, we've modelled ourselves commercially on the notion. We're only financially successful when the customer's service is successful. We charge very specifically for automated resolutions that are completed entirely by AI. But we've retained a hybrid model because we believe the human in the loop remains important."

If a customer comes into an organisation through their chat, WhatsApp or some other digital channel, and the agentic service agent completes the request without human intervention, there's a cost model based on a digital experience. If the interaction goes through a human agent, it's handled through seat-based licensing. Hackney noted that customers are completely empowered to trigger the decision to engage a human agent.

"We will only charge our customers for a successful resolution when we have one. The query needs to be resolved and it needs to not be reopened within three days," she explained.

Zendesk's Resolution Platform moves customer support beyond ticketing toward true resolution. By combining autonomous agents, integrated back‑end connectivity, and continuous learning, it reduces friction, speeds outcomes, and boosts loyalty. The outcome‑based pricing model aligns cost with success, rewarding businesses only when issues close and are not reopened. This approach signals a shift to value‑driven customer service.