Mobile Ad Fraud—Beware! It’s Trending

Last month, when Uber filed a lawsuit against its mobile advertising agency Fetch, it did for marketers globally what the ride-sharing platform did for commuters- transformed the industry by opening it to international challenges, criticism, and scrutiny.

Between 2015 and 2017, Uber paid the agency over $82.5 million for its advertising efforts, of which Uber now claims a sizeable proportion was towards clicks generated through device farms. Clicks were generated on adverts, through multiple devices at a device farm, whose IP addresses were later reset. This gave the impression that downloads were done by multiple real users.

But Uber’s case of fraud is not an isolated one. Another less spoken case is that of consumer goods giant, P&G. The company has recently slashed its digital advertising budgets by over $100 million, on account of online fraud and for being placed next to politically offensive content. The marketing team has chosen to reduce the number of media buying platforms used and are looking to deploy a whitelist which limits ad appearances on only listed websites.

Although Uber and P&G belong to two completely different industries, the problem of digital ad fraud, whether for web or mobile traffic, is common to both.

Mobile Ad Frauds in India

The amount of advertising revenue exhausted globally on fraudulent advertising remains a topic of discussion amongst practitioners. Association For National Advertisers has estimated a global economic loss of $6.5 Billion for 2017 due to fraudulent advertising. Specifically, for India, TUNE’s mobile fraud report states that ad fraud is a rising problem in the country with mobile fraud already being 2.4 times worse than the global average.

India also ranks first for the highest video ad fraud. As the Indian marketing teams start deploying more budgets digitally, the problem of ad fraud is expected to worsen.

“Our experience and growth in the Indian market is evidence enough that the ad fraud industry is advancing year on year,” said Amit Relan, Co-founder of Mfilterit, an Indian ad fraud prevention service by NETAD.

“Bot technology is increasingly mirroring human behaviour making it harder for advertisers to distinguish between real and fraudulent users”. Speaking about Uber, he continued, “Clicks and engagement generated from such device labs appear real, but have no value to brands’ KPIs. Frauds like these are consuming advertisers’ marketing budgets, and putting the entire affiliate marketing industry in jeopardy”.

Kinds of Fraud

With the ever-growing portfolio of the kinds of frauds, Mfilterit’s research has shown that two categories dominate the portfolio for the app market. The first category, device fraud, is the most common. This includes mobile app installs from simulated devices, duplicate users, unauthorised downloads, downloads from incorrect regions, and incorrect telcos. Click fraud is the second category in which fake clicks are generated by ad publishers to capture attribution for installs generated from other sources.

“India ranks amongst top 5 in digital ad fraud and IVT traffic. For online video ad fraud, India is ranked first,” said Manav Sethi, CMO of ALT Balaji, a global video on demand app from India.

“Every marketer’s challenge in the new world order (digital) is ‘identity’ and ‘attribution’”.

Launched in April, ALTBalaji is one of the few advertisers in the country that has pledged to fight fraud in such depth from the inception of their marketing journey. Whilst scaling their user acquisition aggressively in the early months through an attribution platform, they have sought assistance from a neutral third-party Ad fraud prevention service.

The teams have worked together to better optimize 25% of the Ad spend that would have otherwise been paid to malicious publishers.

Prevention

Introducing fraud prevention solutions at an early phase of the company’s marketing activity, is the best strategy to adopt when deciding affiliate marketing budgets. However, a late response to combating fraud is better than no response.

To identify fraud, marketers need to study their affiliate data, and observe any abnormalities. A common rule of thumb remains that if the data is too good to be true, then it probably is. Another clear indication of fraud is stagnant organic traffic or high inorganic conversions. This could signify that organic data is being captured by fraudsters and converted to inorganic. Marketers can also look for repetitive patters such as repeat IP addresses, and browser data to identify activity from same publishers or users.

Advertisers can also track publishers offering incentivised downloads, which often result in poor-quality traffic.

Finally, advertisers can track adverts for their apps on non-play-store APK sites, and identify publishers that are advertising on there, in-turn driving low engagement users.

Whilst all these tips may seem overwhelming for new advertisers, alternatively there are automated tools available to help advertisers with fraud detection. These tools track all parameters of a user’s device and distinguish fraudulent users from real users. Mfilterit’s Ad Fraud detection service is one such tool amongst many, that is based on complex algorithms derived from detailed machine learning. The tool seamlessly integrates with the advertiser’s existing attribution platforms, and collaborates with the advertiser and publishers to reduce levels of fraud.

Whilst the strategy of redressal against fraud is gaining a profound momentum through international organisations such as Trustworthy Accountability Group and GSMA, Indian advertisers have a long way to go before they can achieve a 100% fraud free status. Ad fraud prevention needs to achieve a critical status in the marketing industry, to combat this inherent problem.