For mobile advertisers, the issue of fraud is as up-to-date as ever. Fraudulent activities take up large portions of advertising budgets, stealing away from $16 billion to $19 billion digital ad spend a year, according to different estimations. Overall, 9% of total digital advertising spend is wasted on bots and fake traffic, says Juniper Research.
The issue is evolving: fraud strategies develop and become more sophisticated, even as marketers try to combat them. That is why it is crucial to understand, define and detect fraud, to avoid wasting resources on poor quality traffic.
What is mobile ad fraud?
Mobile ad fraud is a practice of cheating advertisers by fraudulently representing online advertisement events, such as impressions, clicks or installs, in order to generate revenue. Fraud permeates the industry and affects all payment models. Fraudulent activities include:
- serving ads that have zero chance to be viewed by users. This is a way of cheating CPM (cost per mille) payment model.
- serving fake or low-quality clicks, serving motivated traffic instead of unmotivated. Advertisers who work by CPC (cost per click payment) model are at risk.
- serving fake or low quality installs to advertisers who pay per install. This type of fraud harms CPI (cost per install) campaigns.
There are various techniques that fraudsters may apply:
- Manipulating impressions by serving several banners at a time, one on top of another, so only one of the stack is actually visible to the audience while charging all the advertisers for impressions.
- Associating a fake transaction ID with a website that has no affiliation with the app, in that way getting credit for clicks as install source.
- Manipulating post-back event code, so that the system believes the app install happened while it didn’t.
- Utilising bots to install apps, instead of real users.
- Using so-called “human farms”, the groups of people specifically employed to perform fraudulent actions, like installs, in-app actions, purchases, etc.
It is understood that intentional serving of low-quality traffic is also considered fraudulent, including the installs completed by humans, who are not a target audience and are not actually interested in the app. Or advertiser did not express specific interest in this audience. Serving ads to such audience often happen as a result of undisclosed re-brokering happens when there are too many counterparts, and the process of the buying-selling of an offer is so complicated that an advertiser doesn’t know where the install had been performed.
These fraudulent activities can cause financial losses, as well as skew ROI, ARPU, LTV indicators.
How to detect and prevent fraud?
Advertisers and marketers can get ahead of fraudsters by closely monitoring ad campaigns. Eventually, bots will reveal their fraudulent nature, by acting in a way which is not typical for a real user. Some of the features of fraud are the following:
- Unusually high download rate from one source
- The period between click and install is too short
- Identical behavior patterns for the group of installs (install-opening-action with the same time period or same sequence of in-app actions)
- Different geos for click and install or other actions
- Bulks: numerous actions from the same device ID or IP address
- Click IP = Advertiser’s IP
- Other inconsistencies, e.g. the difference between click and install geo, the difference between geo of download and app language settings, high install rate at night, etc.
When advertiser works with dozens of apps, it becomes challenging to detect suspicious activities for each placement. Moreover, only the big data and machine-learning algorithms can identify some of the sophisticated fraud tactics. But it is still essential for advertisers to analyze general trends, and make informed assumptions about the ad campaigns.
Another tactic is to curate and update a list of blacklisted IPs. With the information of suspicious publishers, it is possible to block fraudulent IPs after post-install verification.
Advertisers also can use an anti-fraud solution or connect to a platform that uses anti-fraud. Especially it is relevant for advertisers working with RTB marketplaces and buying traffic by CPM model. Anti-fraud technology can discern fraudulent activities, review impressions and check viewability even before the bid on impression is placed.
How does Clickky fight fraud?
As a user acquisition platform, Clickky takes charge of the quality of traffic delivered to the advertisers. So how do we tackle fraud? We combine Clickky’s internal fraud prevention system with internationally recognized anti-fraud services, such as The Media Trust, Forensiq and FraudScore. It helps us monitor and detect any suspicious or potentially fraudulent activities and block unreliable sources on-the-go, so the advertisers lose as little money as possible.
The present state of the industry calls on advertisers to pay particular attention to combating fraud. The fraudulent activities increase costs of advertising and undermine trust between the partners. Only by addressing this problem directly and allocating sufficient time and resources to fighting fraudulent activities, the advertisers can get ahead of fraudsters and enjoy better results for their ad campaigns.