AI + leadership: the attrition fix for contact centers
Attrition is the silent tax on every support team. You lose your best people, knowledge walks out the door, and the rest carry a heavier load. The fix isn't a single tool - it's AI guidance paired with real leadership.
Customer experience expert Adrian Swinscoe spotlighted how this combo is working in practice after speaking with Olivier Jouve, Chief Product Officer at Genesys. The message is simple: use AI to reduce friction in the work, and use leadership to make the work worth staying for.
How AI guidance actually reduces attrition
Genesys and ServiceNow now connect via an Agent2Agent (A2A) protocol. Genesys AI agents can find the right ServiceNow AI agent, pass front-office context, complete a back-office task, and return results - without passing a caller around or dumping workload on a human.
Results are stacking up. Best Buy Canada, after moving to Genesys Cloud, reported a 20% cut in operating costs, a 19% drop in average handle time, and a 40% reduction in call transfers. Most importantly for teams: real-time guidance and performance support are reducing turnover by 10-30% in some cases.
- Less task-switching and fewer transfers = lower stress.
- Real-time prompts, next-best-actions, and knowledge surfacing = fewer "I'm stuck" moments.
- Consistent quality without micromanagement = more confidence on tough calls.
Curious about the platform behind this? See Genesys for context on the AI guidance stack.
Why leadership and culture still drive retention
Tech won't fix a trust problem. Agents are handling more complex, more emotional conversations - often from home. As Jouve put it: "Imagine if you have 30,000 agents that you have to replace every year. That's crazy."
Sustainable retention needs managers who set clear expectations, protect focus, and give people room to operate. AI can flag burnout and suggest wellness interventions, but humans need the autonomy to act on them.
- Trust: Let agents use judgment on escalations, callbacks, and follow-ups.
- Clarity: Tighten 3-5 quality standards and remove the rest.
- Support: Normalize micro-breaks after heavy calls; coach, don't police.
- Fairness: Balance metrics with call complexity; don't punish the hard calls.
A practical 30/60/90-day plan
- Days 1-30: Baseline attrition, transfer rate, AHT, first-contact resolution, coaching hours. Pilot real-time guidance for the top 5 high-stress intents. Turn on burnout signals (after-call notes sentiment, back-to-back intensity).
- Days 31-60: Roll A2A to 3-5 back-office tasks (refunds, entitlement checks, order status). Update SOPs for "AI completes, agent confirms." Retrain QA to evaluate decisions, not scripts.
- Days 61-90: Tie manager scorecards to coaching quality and voluntary attrition. Expand guidance coverage to 70-80% of volume. Run stay interviews and fix the top 3 causes of churn.
Metrics that matter (and how to use them)
- Voluntary attrition: Track monthly and by cohort (tenure buckets). Look for improvements in guided vs. non-guided queues.
- Transfer rate: If A2A is working, this should fall sharply; re-map intents if it stalls.
- AHT with guardrails: Expect a dip; don't chase seconds at the expense of quality.
- Coaching coverage: Aim for 90%+ monthly with shorter, focused sessions.
- After-call sentiment: Use as an early burnout indicator to trigger breaks or schedule tweaks.
Team practices that keep people
- Start each shift with a 2-minute "what's new" guidance brief.
- Give agents a one-click "I need help" that routes to a live coach or AI co-pilot.
- Publish a "known good" call library with annotated transcripts for tricky scenarios.
- Run weekly skip-levels with 5 fast questions: What's hard? What's unclear? What's dumb? What should we stop? What helped most?
Tooling checklist
- Real-time agent assist (prompts, steps, policy checks)
- AI-to-AI task handoff (A2A) for back-office execution
- Knowledge surfacing with feedback loops
- Quality automation on decisions, not just phrases
- Wellness signals baked into WFM and schedules
Upskill your support team on AI
If your team needs a fast way to get fluent in AI-assisted support, explore practical programs by role here: Complete AI Training - Courses by Job.
The takeaway: let AI handle the grunt work and guide the moment. Let leaders build trust, clarity, and breathing room. Put both in place, and attrition stops being "the cost of doing business" - it becomes a fixable problem.
Your membership also unlocks:
