Fresha's Nova now resolves 80%+ of support tickets with a 4.6/5 CSAT
Fresha announced that its proprietary AI Agent, Nova, is now handling more than 80% of all global support tickets autonomously. It's doing this at scale-thousands of inquiries per day-while maintaining a 4.6/5 customer satisfaction score.
For context, that puts Nova's support experience among the highest-rated in SaaS and technology, not just within beauty and wellness. For support leaders, this is proof that high autonomy and high quality can coexist-if you build the right system.
Why this matters for support leaders
- Hit response and resolution SLAs without inflating headcount.
- 24/7, multilingual coverage becomes practical, not theoretical.
- Quality gets more consistent across regions, tiers, and time zones.
- Agents focus on exceptions and relationship work instead of repetitive tasks.
What Nova actually does
Nova uses advanced AI, machine learning, and natural language processing to deliver instant, contextual answers across Fresha's global network of 140,000+ partner businesses and hundreds of thousands of professionals. By resolving the majority of inbound requests automatically, Nova cuts wait times while preserving accuracy and intent fit.
The result: a support experience that scales seamlessly across markets, business types, and channels-without trading speed for quality. If you need a refresher on how CSAT is typically measured, this performance is genuinely strong.
Security and trust aren't an afterthought
Fresha's AI isn't limited to support. The company runs AI-driven security agents that proactively monitor transactions, detect anomalies, and prevent malicious activity-blocking 99% of fraud attempts across the platform.
That protects both businesses and consumers and reinforces trust across the billions in annual GMV processed through Fresha's marketplace and payments ecosystem.
AI beyond support: operations that move faster
Fresha has embedded AI across onboarding, daily operations, and growth. New businesses can set up profiles in minutes with intelligent onboarding, and professionals can generate full service menus with optimized descriptions, structure, and pricing-removing guesswork.
AI also streamlines appointment management, marketing campaigns, client communications, and revenue optimization. It's a practical blueprint for how an all-in-one platform-booking, payments, subscriptions, marketing, embedded finance, and marketplace discovery-can run with less friction.
Leader takeaway: how to reach 80% autonomous resolution
- Start with your top intents: Map the top 30-50 ticket reasons by volume and value. Prioritize low-risk, high-repeat scenarios.
- Tighten your knowledge base: Write task-first articles with clear steps, policies, and limits. Remove ambiguity and stale content.
- Define guardrails: What can the AI do? What's read-only? What requires human approval? Codify it.
- Build clean escalation paths: Escalate by intent and risk, not frustration. Include structured handoff notes.
- Instrument everything: Track containment, CSAT, first-contact resolution, reopens, and safety events per intent.
- Human-in-the-loop QA: Weekly sampling on critical intents. Close the loop with content and policy fixes.
- Action access: Let the AI perform bounded updates (e.g., scheduling, basic account changes) with reversible steps and audit logs.
Metrics that matter (and keep you honest)
- Autonomous Resolution Rate (ARR): % of tickets fully resolved by AI without handoff.
- CSAT on AI-handled tickets: Track by intent and channel, not just aggregate.
- First-Contact Resolution (FCR): Especially for top intents.
- Escalation Quality: Handoff completeness and time-to-human.
- Safety: False positives/negatives on fraud or policy-sensitive actions.
- Cost per Resolution: Compare AI vs. human across intents.
30-60-90 day rollout playbook
- Days 0-30: Intent analysis, KB cleanup, policy codification, safe tool access, and sandbox testing.
- Days 31-60: Go live on low-risk intents and channels. Daily monitoring, weekly QA, quick-hit content fixes.
- Days 61-90: Expand to medium-risk intents, add more actions with approvals, introduce multilingual coverage, publish runbooks.
What stays human (and should)
- Edge cases with financial discretion, complex policy interpretation, or multi-party coordination.
- High-emotion scenarios, VIP accounts, and regulatory or clinical questions.
- Feedback discovery, product insights, and relationship-building moments.
Quote worth noting
"At Fresha, we've always believed that the best technology should feel effortless. AI Agent Nova isn't just automating support - it's elevating it," said William Zeqiri, Founder and CEO of Fresha. "When over 80% of all support interactions are handled by AI with a satisfaction score that rivals the best human-led teams on the planet, that's not incremental improvement. That's a paradigm shift for our industry."
Bottom line for support teams
High-autonomy support isn't a moonshot. With clear intents, strong guardrails, and tight feedback loops, you can push automation hard without sacrificing CSAT.
Fresha's Nova proves it: faster responses, lower costs, and quality that customers trust. The next step is execution.
Further resources
Your membership also unlocks:
