Success Story
“We had two completely different kinds of customers. Enterprise clients managing million-dollar content programs, and freelancers wondering where their paycheck was. We were routing both into the same queue. Something had to change.”
The Challenge: Two Sides of a Marketplace. One Overwhelmed Support Team.
Contently isn’t a typical SaaS company.
On one side: some of the world’s largest enterprise brands ; financial services giants, healthcare conglomerates, Fortune 500 marketing teams; managing complex content operations across workflows, freelancer networks, and editorial calendars.
On the other: 165,000+ freelance creators expecting fast answers on payment timelines, assignment status, and contract details. Fast meaning now.
When both sides of a marketplace need support, ticket volume doesn’t grow linearly. It multiplies.
For Sarah Mitchell, VP of Customer Experience at Contently, the situation had become structurally unsustainable.
The freelancer side was the pressure valve. Payment status. Assignment updates. Submission confirmations. Queries that were entirely answerable ; if you had the right system to answer them in seconds, not hours.
The enterprise side demanded something different: context, continuity, and confidence that a complex workflow question wouldn’t disappear into a growing queue.
Contently needed a way to serve both, without doubling their support team.
“The repetitive tickets were drowning us. Our best agents were spending half their day answering ‘where is my payment’ while enterprise clients waited on questions that actually needed expert attention. That imbalance wasn’t sustainable.”
Adopting Kayako: Start Small, Scale Fast
Contently started with the highest-volume, lowest-complexity segment: freelancer queries.
Kay; Kayako’s AI support agent; was deployed on the categories that consumed the most tickets but required the least judgment. Payment status. Assignment acceptance. Profile and billing updates. The questions agents had answered thousands of times.
Kayako plugged directly into Contently’s existing helpdesk via API. One agent seat. No platform migration. No disruption to existing workflows. Kay started ingesting tickets within the first hour.
Week one, the support team shadowed Kay’s responses. By week three, they stopped checking. Accuracy was consistent. CSAT didn’t move. The queue got quieter.
That silence earned trust. Then Contently opened the next door.
Implementation Timeline
Month 0 — Freelancer queries go live
Kay deployed for payment status, assignment acceptance, profile & billing updates ; the highest-volume, lowest-complexity ticket categories.
Month 2 — Enterprise expansion
Rolled out to enterprise onboarding FAQs and platform how-to queries. Escalation paths configured by customer tier.
Month 4 — Full-platform deployment
Kay active across both user bases. 68% autonomous resolution locked in. Zero additional headcount despite 40% creator base growth.
“The first week, agents double-checked every Kay response. By week three, they’d stopped ; not because we told them to, but because Kay was consistently right. That’s when we knew we had something real.”
How Kay Worked: Atomic Tools, Not Magic
Kay isn’t a chatbot. It’s an AI employee that operates through composable tools ; reading tickets, querying account data, searching Contently’s knowledge base, composing contextual replies, and closing tickets end-to-end.
| Kayako Capability | What it solved at Contently |
|---|---|
| Intelligent Triage | Auto-classified every inbound ticket by user type (freelancer vs. enterprise), intent, and priority ; before a human agent ever saw it. |
| Autonomous Resolution | Handled all freelancer payment, assignment, and account queries end-to-end without human intervention. |
| Agent Assist | Drafted replies for complex enterprise queries, surfacing relevant knowledge base articles and flagging suggested actions. |
| Escalation Intelligence | When tickets exceeded Kay’s confidence threshold, it compiled a full handoff summary so agents never asked clients to repeat themselves. |
| Self-Updating Knowledge | Auto-generated KB articles from resolved tickets ; turning every resolution into institutional knowledge that compounded over time. |
The result felt like Contently had hired a specialist for every ticket category: one who never slept, never had a bad day, and got smarter every week.
The Results: Support Scaled. Costs Didn’t.
Within 90 days of full deployment, the numbers told the story.
Performance Metrics: Before vs. After Kay
| Metric | Before Kayako | After Kay | Improvement |
|---|---|---|---|
| Avg. First Response Time | 6.4 hours | 91 seconds | ↓ 98% |
| Autonomous Resolution Rate | 0% | 68% | ↑ New capability |
| CSAT Score | 3.7 / 5.0 | 4.6 / 5.0 | ↑ 24% |
| Freelancer Query Handling | Agent-only (manual) | Fully automated by Kay | ↓ 100% agent effort |
| New Hires Needed | 3 agents projected | 0 | ↓ $1.8M avoided |
| Escalation Accuracy | N/A | 94% rated ‘appropriate’ | ↑ First-pass quality |
The $1.8M Math
To absorb Contently’s ticket growth without Kay, the support team would have needed to scale from 12 to 15 agents ; a fully-loaded increase of roughly $390K per year. Over 18 months, and accounting for ramp time, knowledge transfer, and management overhead, the projection reached $1.8M. Kayako’s annual cost came in at less than 9% of that figure. Before factoring in faster resolution times, improved CSAT, or the compound value of Kay’s self-building knowledge base. Growth stopped requiring proportional headcount.
“We used to joke that our support team had the hardest job in the company ; answering the same payment question 200 times a week while also managing a Fortune 500 client’s editorial workflow. Kay took one of those problems completely off the table. I can’t recommend Kayako enough.”
Saw the Results? Now Let’s Talk About Yours.
Kayako resolves up to 80% of inbound tickets autonomously ; without platform migration, without per-seat pricing, and without adding headcount. Kay plugs into your existing helpdesk and starts earning trust from day one.
