Plarium neutralizes 20 million malicious requests each month without blocking legitimate bots

Plarium’s Battle Against Malicious Bot Traffic: A Case Study in Traffic Management

Plarium, a Ukrainian gaming studio renowned for its multiplayer online games, is grappling with a challenge all IT directors know too well: distinguishing harmful automated requests from legitimate human activity in real time. With over 20 million malicious requests filtered monthly, Plarium’s experiences shed light on the complexities of managing web traffic in an environment increasingly susceptible to automated attacks.

Boasting several hundred million registered accounts, Plarium’s platforms are home to high-traffic authentication endpoints, continuous registration flows, and a reliance on analytical data to guide acquisition budgets. This unique combination renders it a prime target for three types of automated attacks: credential stuffing at login pages, mass creation of fake accounts, and artificial traffic generation aimed at distorting marketing metrics. The implications of these challenges extend well beyond gaming, affecting sectors such as e-commerce, financial services, SaaS platforms, and digital utilities.

Traditional network analysis solutions typically assess malicious traffic based on network layer indicators: request volume from a single IP address, login attempt frequencies, and the geographic location of traffic sources. However, these methods often fall short against modern attacks that distribute traffic across residential proxies, impersonate legitimate user behavior patterns, and maintain unit volumes low enough to avoid triggering conventional thresholds. The result is a dual challenge: under-blocking allows distributed attacks to slip through, while over-blocking hampers legitimate users sharing the same infrastructure.

An Unprecedented 35-Fold Surge in Traffic Handled Seamlessly

Before adopting DataDome, Plarium evaluated several vendors whose strategies relied heavily on network indicators. As Plarium’s Chief Information Security Officer Oren Zenescu expressed, “We don’t want to block all bots, just the malicious ones. With DataDome’s client and server signals, we achieve the necessary precision without compromising the player experience.” This distinction is critical: search engines, monitoring tools, and data aggregators comprise a portion of non-human traffic that the operator strategically needs to preserve. A blanket approach to blocking automation could negatively impact SEO visibility and platform observability.

An illuminating incident during this deployment underscores the scale of the challenges at hand. During a documented attack, Plarium experienced a staggering 35-fold increase in traffic across its endpoints, with real-time analysis revealing that 99% of this surge was automated and malicious. Remarkably, detection and blocking took place without interrupting service for users, who experienced no noticeable degradation. This outcome highlights the limitations of conventional volume-threshold approaches: blocking triggered beyond a certain traffic rate inevitably disrupts legitimate users during organic spikes, such as content launches or seasonal events.

Safeguarding Analytical Data: A Structural Advantage

The Plarium case study also highlights an often-overlooked collateral benefit of bot protection: the integrity of analytical data. The marketing team runs multi-channel acquisition campaigns with budgets allocated based on measured performance. When bot traffic mingles with legitimate traffic in attribution tools, it distorts registration metrics and renders channel evaluations unreliable. “Return on investment is calculated based on human clicks, not automated ones,” Zenescu notes. Intent-based detection restores reliability to behavioral analysis cohorts and acquisition dashboards, directly influencing the quality of marketing investment decisions.

This friction point can be applied across various operators utilizing registration forms or conversion pages exposed to external traffic and relying on analytical data for budgetary arbitrage. Bots testing stolen credentials on login pages generate failure events that clutter dashboards and can obscure warning signals. Application traffic protection and analytical data governance are no longer separate concerns when automated traffic reaches the volumes documented by Plarium.

In summary, Plarium’s proactive measures against automated attacks not only safeguard its gaming infrastructure but also enhance its analytical reliability, serving as a valuable lesson for industries facing similar challenges.

Scroll to Top