Harnessing AI and Automation: How Bytro Labs Manages 17K Monthly Tickets with a Lean Team
17K
Monthly Issues
2.5
Agents
99%
Automation Rate
20
Days to Implement
Managing a volume of 17,000 tickets per month with a lean team of just 2.5 agents is no small feat, but our system has proven remarkably effective. Automation has been a game-changer, enabling us to handle this workload efficiently and maintain high-quality support. I’m impressed with how well this approach works and confident we couldn’t achieve these outcomes without leveraging advanced automation.
The games have over 15 million players and are available in over 12 languages. Despite handling nearly 17k support tickets per month, Conor’s powerhouse team of 2.5 agents manages the load through strategic AI and automation. A key part of this strategy includes proactively addressing toxicity and cheating—issues that can quickly escalate. By leveraging Smart Intents, the team identifies and handles nearly 140% of support volume related to toxicity, using dedicated moderators to ensure a safe and fair gaming environment.
Dive into this case study to uncover Conor’s best practices for rolling out an in-game support system to handle a high volume of issues with a lean team.
The Challenge
Previously, we relied on contact forms and emails, with volunteers managing the issues. Tracking issues based on specific locations, game states, or trends was extremely difficult. Handling a high volume of cases became nearly impossible without app-provided context and metadata. We needed a more consistent way to identify who and where the players were, what they were doing, and how they were doing it.
The old system also had significant flaws, such as disappearing tickets and no submission tracking. Multi-account users faced challenges when opening tickets via email, adding to the complexity.
Additionally, the lack of tracking made it hard to monitor performance, as there were no CSAT metrics. Generating reports was tedious, and the results weren’t actionable.
The Solution
Self-service As The First Layer of Defense
We ensure that tickets that don’t need an agent’s attention don’t reach an agent. Our first line of defense consists of self-help resources: FAQs and tools that empower players to resolve issues directly. In the next layer, we introduce extensive automation, anticipating players’ questions and addressing them in advance. We convert repeated inquiries into FAQs and remove any unnecessary steps. By the time an issue reaches a human agent, it’s because it genuinely requires human intervention.
Combining Skilled Humans with Advanced Automation
By designing seamless automation and processes, we ensure that our skilled agents focus on more complex issues, keeping their work engaging and adding greater value to the players. Our team of agents is fantastic, and their exceptional skills ensure everything runs smoothly. But it’s our automation that truly elevates our efficiency.
Enhancing the Experience by Leveraging AI Helpshift’s AI-powered Answers is incredibly valuable. It allows players to get the information they need without making an extra click to access the full FAQ, which is where players often decide to bypass self-service and request a human agent. With the AI-generated summary based on resources we’ve created, we can trust that the information provided is accurate and relevant to our game, ensuring consistency and reliability in player support.
Intent Detected for 93% of the Issues by Smart Intents
We’ve integrated Smart Intents into every interaction because understanding player issues is crucial. The system accurately understands incoming player queries and automatically routes them to the most appropriate workflow for resolution.
The Results
Support Organization
63.7% deflection rate
99% automation rate
$850K savings per year
Player Experience
74% of issues come from in-app, 25% via web chat, and only 1% via emails
Helpshift successfully detects intent for 93% of issues
4m TTFR (Time to First Response)
Addressing Toxicity and Cheating Through Smart Intents
Smart Intents allows us to identify a significant volume of toxicity in player issues—nearly 40% of the total volume, with 21.3% related to toxicity and 18.8% to cheating, primarily reported by players. By offering a structured process for reporting toxicity and cheating directly via Helpshift, the support team can direct players to moderators specifically trained to handle these sensitive issues. We review the analytics to identify which features are being abused, how they’re being abused, and the types of comments we’re handling. This includes issues like extremism, general vulgarity, or excessive profanity. We can take appropriate action once we know what we’re dealing with. The process ensures we can address and understand harmful behavior in our games, protect player well-being, and maintain a safe gaming experience for our players.
Top 10 Intents Groups by Volume
Top 10 Intents by Issue Volume
“With Smart Intents, Helpshift helps eliminate blind spots when addressing toxicity.” – Conor McGinley, Customer Support Manager at Bytro Labs.
