How Pocketmath fought malicious actors in 2016

Posted by Karthic Nithyanandan on Feb 20, 2017 4:34:50 PM

Topics: Industry News & Trends, Company News

2016 saw an escalation in the war against malicious actors at Pocketmath, and we’d like to share our battle experiences, as well as our shared responsibility in making the mobile marketplace safer.


How it all started?

You might all agree that the term malicious is extremely general, and it was really through experience that Pocketmath teased out the actual meaning.


Bad Ads.png


A common practice employed by the malicious actors in 2016 was that they were constantly hitting the growth DSPs—and Pocketmath being one of the fastest evolving DSPs was no exception to it. The surge in such activity started In April 2016, when we started receiving serious complaints from our supply partners about ads delivered using our platform. We were quite taken aback by this feedback: from the start, we know we are in an industry where a creative is supposed to tell a positive story about a brand, product, or service. And Pocketmath serves ads to millions of users. Being well aware of our responsibility, we’ve always worked to qualify and ensure proper standards from our customers. However, after being questioned about the quality of our demand, we decided to take vigorous measures to protect the publisher space and the safety of end users.


It was time to make some changes!


What did we change?

  • Vetting Technology: Since the beginning of the business, we have had our in-house vetting technology complementing and helping us to drive the creative-approval process. During this journey, we recognized the importance of industry standards and benchmarks for monitoring ad quality and, therefore, invested in high-end technology and resources to enhance our tools. We also partnered with players such as The Media Trust and GeoEdge.
  • Creative Review Processes and Timelines: To be as fast as the RTB industry, we believed in real-time creative approvals. We soon came to realize that this is a double-edged sword. Consequently, we introduced creative-approval timelines so as to audit creatives based on their formats. The extended timelines gave a fair amount of time for our tools to analyze and escalate unusual creative behavioral patterns. We also paved the way for pre-vetting and post vetting phases, introducing different scanning methodologies.
  • The Human Touch: Although we have our machine-learning automation scans, we believe humans are much better at making complex judgements. To these ends, we employed a human verification team to run live tests, examining and verifying the ad-quality standards so as to replicate an end user’s experience.
  • Recognized Partners: Having studied industry trends, we noted that custom creative tags or creative tags developed in-house are often deemed to be more vulnerable. As a result, we gave more leeway to recognized vendor tags, such as DoubleClick and Sizmek, and also began to educate our clients on adopting such recognized formats. 
  • Creative Policies: The biggest challenge in the programmatic world is that different partners have different policies; it’s never easy to come up with a single common solution. Therefore, we enhanced our policies and created a centralized set of rules that were incorporated into our system. We also maintained constant communication with our ad exchange partners to help them keep abreast with their policies. For example, in line with a few industry updates, we now allow creatives with iFrames on certain SSPs.
  • Customer Qualification: Of course, the low-hanging fruit was to improve our customer qualification process, and we made a significant change in revamping practices in this regard.
  • Report to Track Creative Inconsistency: When submitted creatives are flagged and rejected, one of the common statements we often hear is "Oh! DSP X allows these tags. So why don’t you?”  To ensure that our clients are more confident in our service, we have also invested more time and resources in building out reports to track creative inconsistencies.




The consequence

In 2016, the creative influx rate doubled as compared to 2015. Pocketmath invested time, resources, and technology and stood as a foe against the malicious actors.


Auto-redirects was one of the top flagged violations. The most popular and polluted form of violations, these are generally created by sophisticated hackers. The primary focus of auto-redirects is to drive fake conversions and clicks, so as to attain the required performance marketing objectives. Such creatives reveal their actual behavior only when certain conditions are met and ultimately spoils and interferes with the end user experience—thereby being a primary cause of violations from SSPs.


  • In our machine-vetting stage (before approving a creative to compete), 36% of ads scanned violated our advertising policies in 2016.
  • Overall, we reassessed all our customer accounts and decided to terminate 200 accounts in order to improve the quality of our demand.


Applying what we learned: looking ahead to 2017

While malicious actors get smarter day by day, we are trying to be two steps ahead in prioritizing brand-safety measures.


  • Recognizing the importance of defending supply-demand harmony is the key factor.
  • Transparency is another key factor. Clients become more agreeable and understanding when we address the “why” questions with supporting reports and workarounds.


We will continue to stay committed in keeping the mobile marketplace safe for both our valued clients and their consumers. 



Would you like to run a mobile advertising campaign safely?




Subscribe to Email Updates

Popular Posts

Recent Posts