“With all of the data that is available, it has become an impossible task to track everything regularly, especially for secondary KPIs and sub-metrics.”
GCash Reclaims Two Working Days a Week With AppsFlyer Agent Hub
2 days
back every week on manual monitoring40%
of a full-time analyst's yearly workload back GCash slashed anomaly-detection time from weeks to days and reclaimed two working days a week with AppsFlyer Agent Hub.
• GCash’s performance marketing team manages dozens of ad networks and a high volume of in-app events, but manual, spreadsheet-based monitoring meant only a handful of top networks ever got a real deep-dive, while secondary KPIs and sub-metrics went largely unwatched.
• That gap let budget-gaming networks slip through, and it nearly cost the team a high-value partner that looked weak on topline numbers but was quietly driving strong secondary conversions.
• With AppsFlyer Agent Hub, including an agent the team built to map its own event taxonomy and funnel, GCash now measures 100% of its network portfolio continuously, catches anomalies in days instead of weeks, and has reclaimed two working days a week.
Background
GCash is the Philippines’ leading mobile wallet and super app, processing a large volume of in-app events across a broad portfolio of financial products. Its performance marketing team covers the full stack, operations, campaign management, data, strategy, and technology, for user acquisition across dozens of ad networks at once. That scale is the whole story. The more products and campaigns run at the same time, the harder it becomes to know, at any given moment, what is actually working and where.
Challenge
Deep-dives only reached a handful of networks
Before Agent Hub, monitoring campaign performance meant checking topline metrics and headline KPIs by hand. With GCash running campaigns across a large roster of ad networks, the team could only thoroughly deep-dive a select few each week. The rest went largely unchecked, as Kimberly Beatriz Ong, Performance Marketing Technology and Transformation at GCash, described it:
“With all of the data that is available, it has become an impossible task to track everything regularly, especially for secondary KPIs and sub-metrics.”
Secondary KPIs and sub-metrics went unseen
As a super app, GCash captures a large number of distinct in-app events beyond the primary conversion. Regular monitoring rarely reached past the main KPIs, so sub-metrics and secondary events, the places where early warning signs tend to show up first, were effectively invisible until a campaign cycle had already ended. That blind spot let some networks game the system and run up wasted spend, while other genuinely high-performing partners sat unnoticed and under-scaled.
Solution
Earning trust before automating action
When GCash first deployed Agent Hub, the team manually validated every agent output before acting on it, a deliberate human-in-the-loop step to confirm the AI’s read on the data matched their own. That habit stuck. Today, the team still cross-checks AI insights before making high-impact campaign changes, even as day-to-day monitoring has become automated.
GCash built its own agent to map its full user journey
Every business’s funnel is built differently, out of its own mix of in-app events, KPIs, and definitions of what counts as a drop-off. Agent Hub’s Create an Agent capability is designed for that reality: it lets teams build on top of the ready-made agents and configure their own event-to-KPI mapping, tailored to how their business actually works, rather than a one-size-fits-all default. GCash used it to create an agent that lays out their complete user journey end to end, stitching together the specific in-app events that make up their funnel so the team can see exactly where users are dropping off, instead of inferring it from topline numbers alone. As Ong described it:
“I was able to paint out the full user journey for the agent, helping us to quickly understand if there are any alarming drop-offs within the funnel.”
“I was able to paint out the full user journey for the agent, helping us to quickly understand if there are any alarming drop-offs within the funnel.”
Results
Every network, measured continuously
Agent Hub now runs continuous, `automated monitoring across 100% of GCash’s active networks and sub-metrics, not just the handful the team could reach manually before. Anomalies and publisher discrepancies that used to take weeks of post-campaign analysis to surface now get caught in days. As Ong put it:
“Agent Hub significantly accelerated our issue detection timeline, reducing the time it takes to catch performance anomalies or publisher discrepancies from weeks to days.”
“Agent Hub significantly accelerated our issue detection timeline, reducing the time it takes to catch performance anomalies or publisher discrepancies from weeks to days.”
That automated cross-referencing alone has freed up two working days a week across teams, roughly 40% of a full-time analyst’s yearly workload, while protecting budget by flagging underperforming or fraudulent network activity instantly.
A partner that would have been cut, and became a growth channel instead
The clearest proof the second blind spot is closing: an automated alert flagged an anomaly on a partner network that had been underperforming on GCash’s main KPIs, one the team’s standard manual protocol would likely have paused. Digging into the sub-metrics Agent Hub surfaced, the team found the partner was actually delivering high-quality, high-converting users against a critical secondary in-app event. Instead of cutting it, GCash retargeted and scaled the partner’s budget, turning a network that looked like a loss into a key growth channel.
Looking Ahead
GCash is not stopping at detection. The team’s next step is refining the agent they built to connect in-app funnel events directly to real-time budget optimization, moving from an agent that flags what happened to one that recommends what to do next, including draft budget reallocations across networks. As Ong put it:
“The single biggest change is shifting our team from reactive reporting to proactive optimization. Instead of spending hours digging through spreadsheet rows to find out what went wrong after the budget was spent, Agent Hub brings actionable insights straight to us in real time, allowing us to protect budget and capture growth opportunities immediately.”
“The single biggest change is shifting our team from reactive reporting to proactive optimization. Instead of spending hours digging through spreadsheet rows to find out what went wrong after the budget was spent, Agent Hub brings actionable insights straight to us in real time, allowing us to protect budget and capture growth opportunities immediately.”
