The Dirty Secret of Mobile Ads: How One Bad Creative Can Kill Your Retention
Alex Yerukhimovich, GM of AppHarbr, is blunt about the stakes: one bad ad isn’t just annoying - it’s a revenue leak. From malvertising and forced redirects to deceptive UX tricks, low-quality creatives silently erode retention, trust, and lifetime value. AppHarbr’s approach flips the model, blocking harmful ads before they ever render - forcing a shift from short-term eCPM chasing to sustainable monetization.
What specific criteria and technical benchmarks does AppHarbr use to classify an ad network as “verified” for ad quality and safety?
User Safety
Malicious advertising (malvertising) that harms or compromises users, including malware, phishing, scams, forced redirects, and unsafe journeys leading to fraud. Top Performer status is awarded only to ad networks maintaining a malvertising rate below 1%.
User Experience
Disruptive ad behavior that degrades the app experience across two key areas:
- Unskippable Ads – Ads that exceed app-defined time limits (30 seconds for rewarded, 15 seconds for interstitial), lack a clear skip option, or present broken or unclear close functionality.
- Auto-Play Sound – Ads that play audio automatically, override mute settings, or trigger unexpected sound.
Ad Content
Brand safety and policy compliance violations, including explicit or offensive content, weapons, illegal drugs, or alcohol-related creatives. Top Performer status is awarded only to networks maintaining a violation rate below 0.5%.
How does AppHarbr technically detect and quantify low-quality ad behaviors (e.g., auto-redirects, disruptive interstitials, malicious scripts)?
The AppHarbr SDK is embedded directly within the mobile application, providing client-side enforcement through real-time runtime monitoring of the in-app ad environment.
It performs dynamic analysis of both creative code content and ad behavior, including associated landing pages. AppHarbr’s decisioning layer evaluates creatives against behavioral risk models in real time. If a violation is detected, the SDK prevents the creative from being rendered and replaces it before it is presented to the user ensuring protection occurs pre-impression.
Based on your data, how does ad quality correlate with KPIs such as retention, LTV, ARPDAU, and session length?
Ad quality issues such as auto-redirects that force users out of sessions or offensive ad content that damages trust have a direct impact on user engagement. These disruptions frustrate users, increase churn, and can ultimately lead to uninstalls.
Strong ad quality can uplift these KPIs, but the actual values vary greatly between different apps based on numerous factors including category, geos, ad networks, ad formats and more. Specifically with ARPDAU, for most gaming customers, AppHarbr’s ad quality control normally manage to deliver a 5%-15% uplift.
Do developers typically see short-term revenue fluctuations when removing low-quality networks, and how should they evaluate long-term yield versus immediate eCPM?
Most publishers are well aware of this major challenge - ad networks prioritize max immediate revenue in the form of the highest bidding ad (which is often times also the most intrusive), while ignoring the retention tax that the publisher may pay as a result, which leads to an overall negative ROI, compared to slightly lower-paying but higher-quality ads. AppHarbr solves this challenge by allowing publishers to accept only adequate ads even from low-quality networks without having to remove them altogether. It's a win-win - you get only the highest-paying ads that are good for your business, but can get the next highest-paying one with good quality in case the highest paying one ends up losing money for you due to poor quality
Without breaching confidentiality, what characteristics distinguish the highest-performing verified networks in terms of quality and monetization outcomes?
Top Performer status is achieved by networks that maintain a malvertising rate below 1%, actively blocking malware, phishing, scams, forced redirects, and unsafe user journeys that can lead to fraud. This strong security foundation protects both users and publisher revenue.
Beyond security, these networks demonstrate strict user experience discipline. They avoid disruptive behaviors such as unskippable ads that exceed app-defined time limits (30 seconds for rewarded, 15 seconds for interstitial), broken or misleading close buttons, and auto-play audio that overrides mute settings. Protecting the user experience directly supports retention and long-term revenue growth.
Finally, they uphold high brand safety standards, maintaining a policy violation rate below 0.5% by preventing explicit, offensive, or otherwise non-compliant content from reaching users.
What early warning indicators should developers monitor that suggest ad quality is harming user experience before revenue metrics decline?
Unusually long average display time (in full screen ads), inflated CTR (that suggests many accidental clicks while trying to close hard-to-close ads), increased churn among users who saw a specific ad or ads from a certain ad network, and bad reviews or support tickets that complain about poor ad quality
How does implementing ad quality controls at different stages of the programmatic stack (pre-bid filtering vs. post-bid monitoring) impact monetization efficiency?
(you mean pre-render rather than pre-bid, as before the bid there's no auction winner and ad response to monitor). Pre-render offers the significant advantage of preventing the harmful ad from being displayed to users altogether, and this avoiding any negative consequences whatsoever. Alternatively, identifying an ad only after it has been served is a reactive ad quality measure. While this post-serve detection is commonly offered by network mediation platforms, it does not prevent the ad from being shown to the user in the first place.
How is AppHarbr evolving its product to address increasing regulatory scrutiny, brand safety requirements, and fraud prevention within in-app monetization ecosystems?
AppHarbr already has the capabilities required to accurately detect ads that violate regulations and policies, whether through prohibited content or unauthorized actions taken without user consent. By customizing each publisher’s ad policy settings to align with the specific regulations and internal guidelines they are subject to, we help ensure full compliance while protecting the user experience.









