Head of Content and Mobile Insights, AppsFlyer
How do you make sure that only real people see your ads? According to a leading source of ad tech reporting, ad fraud costs the industry $8.2 billion a year in the US alone. As advertisers move to digital channels, how should you protect your brand and seek better efficiency and ROI for your ad dollars? We sat down with Shani Rosenfelder, Head of Content and Mobile Insights at AppsFlyer, to discuss the state of ad fraud and the major challenges faced by mobile advertising companies.
Tell us about your role at AppsFlyer and the team/technology you handle.
I am the head of content and mobile insights. My team’s job is to create meaningful, value-driven content about mobile marketing, analytics and attribution. This includes different types of content such as blogs, guides, presentations, webinars and most importantly data studies. Because of our role in the industry as a trusted third-party attribution platform and due to our unrivaled scale, our data benchmarks generate significant impact, both for our clients and partners and for us. It is therefore one of the main pillars of AppsFlyer’s marketing activities.
What is the ‘State of ad fraud reporting for mobile apps’ in 2018?
Ad fraud reporting is just beginning to come of age. A good ad fraud reporting solution will (1) block known fraud, (2) report on what was blocked and what has been saved, (3) allow the marketer to dive deeper into their fraud metrics, (4) provide the marketer with the ability to customize their fraud terms and (5) provide transparent, real-time postbacks to ad networks. Anti-fraud solutions have become a very hot space. However, the quality of these solutions, including reporting shows lots of room for growth and improvement.
Other than bots in ad fraud, what are the other major threats to mobile marketing companies?
Mobile marketers face an exciting challenge: virtually unlimited opportunity. There are so many solutions and providers, opportunities for growth and scale, improving user experience, LTV, ROI and more. As a result of this abundance of options, many marketers find themselves stuck in analysis paralysis. My advice to marketers is to find the optimal mix of quality and scale, tried and true solutions and exciting new opportunities. If you aren’t testing something new, you likely aren’t recognizing your full marketing potential.
What regulations are in place to defend brands and customers, respectively, from ad fraud?
While there are regulations against fraud, including device farms, in many regions, enforcement of anti-fraud standards falls to the advertiser, the ad networks and any third-parties they engage. This is why we encourage advertisers and networks to clearly delineate fraud terms in their agreements.
Define fraudulent installs. How does AppsFlyer enable mobile app marketers to prevent this malpractice?
Fraudulent installs are installs that did not occur as reported to the advertiser or their measurement partner. Install fraud can come from malware on a real user’s device, bots residing on servers or device farms (physical locations full of real devices).
What is the most sophisticated tactic to overcome fraud? How does technology help measure legitimate traffic?
Mobile fraud has become an increasingly dangerous game of cat and mouse. Fraudsters are continuously improving their ability to mimic legitimate traffic while adapting to new anti-fraud measures much faster than before. Identifying fraudulent activity and validating a new fraud signature requires a massive amount of data and advanced machine learning capabilities that can actually pinpoint malicious and suspicious patterns in the data. Ultimately, the technology separates legitimate traffic from illegitimate traffic.
What is your 5-point-strategy for marketing teams to tackle the threat of ad fraud?
1) Understand the dynamics of fraud: Mobile ad fraud comes in waves and hits practically every app that invests in user acquisition (we did not find any correlation between an app’s audience size and the rate of fraud). As a result, more and more marketers are now investing in fraud solutions as an insurance policy. This proactive investment incrementally improves their performance data integrity today and provides the coverage needed when the next big wave hits.
2) Focus on blocking fraud in real-time, but remember not all fraud can be blocked in real-time: It goes without saying that you would like to block 100% of fraud in real-time. But know this: there is also a significant risk in overzealous blocking. It can damage your relationships with publishers, impact ad network revenues and compromise data accuracy.
Because mobile is highly fragmented, there are numerous potential points of failure, which could set off false fraud alarms. Unsupervised machine learning is highly effective for detecting anomalies, but labeling every anomaly as fraud would be dangerous and misleading. Even a server delay or an external API bug can result in abnormal activity, but in this case, it is clear that no fraud has been committed.
Knowing when and what to block in real-time without compromising data quality and integrity is of the utmost importance. It requires extreme care and precision, significant R&D resources and a massive amount of fresh data.
3) Update your SDKs to the latest versions: Running the most recent SDK version ensures that you have the latest security updates.
4) Set your fraud terms with every media source: Do this before campaigns begin to avoid the headache of reconciliation negotiations.
5) Never rest on your laurels. Mobile ad fraud is extremely dynamic. It is vital to keep fraud top of mind at all times, even if you experience reduced fraud rates. That is exactly when a new type of fraud is likely to surface, inflicting damage before it is noticed.
Thanks for chatting with us, Shani.
Stay tuned for more insights on marketing technologies. To participate in our Tech Bytes program, email us at firstname.lastname@example.org