As the iOS and Android app business grows, we are seeing more and more bad actors attempt to perpetrate advertising fraud on our industry.
This series of Singular blog posts is designed to provide brief tutorials more mobile device app advertisers on some of the key ways that advertising fraud is committed, as well as provide advice on how to reduce or mitigate the risk to your app business.
App Install Farms and App Store Rankings
“Farms” have been around almost as long as the digital industry. In the PC-centric Internet era, “click farms” were used to manipulate site rankings and affiliate marketing programs by artificially increasing click counts. In fact, click farms were for a time so prevalent that spoof sites like this grew enormously popular as a reaction. As mobile and mobile apps have come to dominate the digital space, a different class of “farm” has emerged in a big way: app install farms.
An app install farm uses a team of people (or technology) to install, launch, and then uninstall apps from devices. The process is then repeated by the farm employee so that it appears that an app is being installed by many people when the reality is that each install is fraudulent.
Download the Singular Fraud Index
SIngular has just released its 2017 Singular Fraud Index, which reveals the mobile industry’s Most Secure Ad Networks. The Index is the first study of its kind to examine fraud data collected from multiple fraud prevention solutions, each with its own set of proprietary detection methods. Drawing on this unique dataset, the study exposes the effectiveness of today’s most common anti-fraud mechanisms as well as the 25 ad networks driving the lowest rates of fraud.
Get your free copy here.
Early on, app install farms were knowingly used by some unscrupulous app publishers to manipulate rankings on the Apple App Store and Google Play. By radically increasing the install count for an app, an app install farm could help an app secure a top ten ranking. Top ranking g radically elevates the profile of an app and often drives many (legitimate) installs because consumers assume the app is strong.
Over time, the app stores caught on to this flagrant attempt to manipulate their data, so they adjusted their algorithms to reward brands that drive persistent installs and penalize brands that exhibit high immediate uninstall rates.
App Install Farms and Advertising Fraud
Today’s app store farms often help bad actor media companies artificially increase the number of ad-driven installs for a brand. Here’s how it works in 5 steps:
- The person first clicks on an ad for an app on each of the N devices s/he manipulates
- They then install the app on one of the app stores
- Then, when the app is downloaded, they launch the app to register the install with the brand or attribution provider
- They then uninstall the app
- Once uninstalled, they click on another ad and begin the process again
The team at a farm would be paid to click ads, then install an app and launch it, so that the media company could register a new install for payment under cost-per-install media schemes.
Driving false installs in this manner can be very profitable for the farmer/fraudster because Android and iPhone app publishers are often invest a great deal of money for installs. Industry estimates of the average cost of an install vary, but most estimates place the worldwide average at around $2. If an app farm worker could operate 100 devices and drive 25 installs per day per device, that means they could generate $5,000 per day for the farm.
Farm-based artificial installs have no value to an advertiser because they don’t represent the actions of consumers, yet they are likely to cost just as much as legitimate installs. That’s why it is so important to detect and prevent them.
App Install Farms Around the World
When we use the term app install farm, we are usually referring to an arrangement where people are used to install, launch, and uninstall apps. Because this is a labor-intensive job, but requires no formal education, such farms are far more likely to exist in developing countries than in the US or the European Union. However, some farms use technology to drive the installs rather than manual labor. Thusm these criminal enterprises exist all over the world.
App Install Farms and the “Ad Fraud Arms Race”
Battling against fraud is a constantly changing proposition because every time legitimate businesses find a way to detect and prevent fraud, fraudsters change their methods. For example, the classic app install farm formula – of rapidly installing, then launching, then uninstalling, and then repeating the process – meant that many installs were being credited to a single device advertising ID (IDFA or Google Advertising ID.)
As advertisers began to block large numbers of installs from the same device advertising ID, fraudsters simply added an ID reset to their workflow. Now savvy anti-fraud actors must monitor their data for lots of new device IDs in addition to multiple installs on a single ID.
It is important to be vigilant as regards monitoring for signs of advertising fraud. Here are a few strategies to help you do just that:
- Anti-Fraud Tools: Some attribution and analytics suites offer tools to help advertisers identify and prevent fraud. SIngular, for example, automatically offers many protections. Such tools may use signals like IP addresses, click and install pattern detection, and activity monitoring to pinpoint campaigns, partners and buying models that are driving suspicious app installs.
- Common Sense: A deal that sounds too good to be true is likely to result in low-quality app installs.
- Focusing Resources on Trusted Partners: Large or niche vertical media companies are more likely to have the scale and resources to detect and prevent fraud. Further, properties like social networks can leverage user account information to help ensure that installs come from legitimate people.
- Leveraging Retention and Uninstall Data: By comparing the set of user traffic attracted by different media companies, brands can learn a lot about user quality. Low user retention or high uninstall rates increasingly are seen as signals of possible fraudulent activity.
- Use ROI Analytics as a Primary KPI:. Arguably, the best KPI to monitor and prevent fraud is ROI, or revenue divided by cost, as it is particularly difficult for fraudsters to simulate a sale, especially because most marketing analytics platforms — including Singular — verify in-app purchases with App Stores.
Fighting advertising fraud isn’t easy, which is why so many companies turn to companies like Singular for help.
Singular and Ad Fraud
Singular offers an industry-leading fraud solutions that you can learn more about right here. For a capsule summary of some of the steps we take to detect and prevent fraud for our clients, read on.
At Singular, our unprecedented visibility into BOTH the ad performance and spend sides of the install category gives us unique advantages in detecting and protecting clients from fraud. Our fraud detection and fraud prevention technologies are a mix of the best known techniques today as well unique proprietary techniques only we can offer.
Singular monitors the flow of data into our platform, looking for signs of fraudulent activity such as illegitimate networks, IP addresses, devices, mismatches in targeting, and more. Further, Singular verifies every in-app purchase (IAP) with the appropriate app store to ensure that you recognize revenue from every reported purchase.
Singular’s New Anti-Fraud Improvements