Go Mobile continues to deal with fraud in digital advertising. Read the new article about metrics to identify ad fraud, benchmarks of different platforms, and the capabilities of popular anti-fraud systems.
This article is the second part of our material about ad fraud. It is dedicated to the methods of fighting fake traffic and the possibilities of anti-fraud solutions.
There are three basic rules for fighting ad fraud:
- Use specific metrics to identify fake traffic.
- Generate benchmarks and track anomalies.
- Use anti-fraud systems (the most important one!).
Follow the link if you missed the first part of the material about the types and mechanisms of ad fraud.
Metrics
To track fraud, there are specific metrics that need to be taken into account when optimizing advertising campaigns.
Key metrics for display and web advertising:
- Bounce Rate — the percentage of users who left the site/app on the login page or viewed no more than one page;
- The proportion of undefined devices;
- Share of non-targeted geo;
- Robotness — the share of bots in traffic;
- Empty Referrer;
- The adequacy of the distribution of sessions by the hour.
Key metrics for performance advertising:
- Share of new devices.
- Click to Install time (CTIT) or Time to Install (TTI) is the length of time from the click on the ad to the first opening of the app.
- Retention Rate (RR) is the percentage of users who returned to the app.
- Churn Rate is an indicator of the user’s outflow from the app.
- Uninstall Rate is the percentage of app deletions from all installs for a certain period.
Benchmarks
The benchmarks for ad fraud are different for each app or website, since the indicators are influenced by the subject of the business, geography, and features of platforms where you are promoted. But first, you can focus on benchmarks in different sources.
We recommend promoting not only in «safe» sources. On platforms with a relatively high proportion of ad fraud, there is high-quality traffic as well. The main thing is to optimize campaigns not only according to KPIs but also according to the data of the anti-fraud system.
Anti-fraud systems
One of the main services of our agency is media buying. We discuss with each client the harm that ad fraud can bring and the importance of using anti-fraud systems.
Go Mobile strongly recommends using an anti-fraud system, as it is often impossible to recognize low-quality traffic without it.
Anti-fraud systems can be divided into two groups:
Anti-fraud systems embedded in tracking systems
Many tracking systems have built-in ad fraud protection. The most often used anti-fraud solutions are from Adjust, AppsFlyer and Kochava.
Independent anti-fraud systems
Independent technologies are mostly more specialized — they can track rarer types of fraud, as well as fake traffic on the web. Anti-fraud systems of mobile trackers cannot implement such functionality.
How anti-fraud systems work
Globally, anti-fraud systems can operate in 3 modes:
Show reports on the probability of fake traffic
Report generation is the basic function of all anti-fraud systems. With their help, you can track suspicious activities and make decisions about how to block them manually. An important point: not all systems explain why they block some events. Argumentation is provided, for example, by FraudScore and AppsFlyer.
Evaluate suspicious actions in real time
Many anti-fraud systems can block fraud installs and clicks, and the actions of bots in real time — using machine learning models or pixels. For example, such functions are offered by the tracking systems Adjust and AppsFlyer, as well as an independent Fraudlogix system.
Evaluate suspicious installs when analyzing post-install indicators
When analyzing post-install indicators, anti-fraud systems detect and block fake traffic after performing deeper actions than the first touch. As example, by studying the activity of the listed users in the app or on the site. This mode of operation is available only in tracking systems.
How to choose an anti-fraud solution
To choose an anti-fraud system suitable for your needs, you need to take into account 4 factors:
Factor #1. Web or mobile
Only independent anti-fraud systems can track fraud on the web. And tracking systems themselves can fight fraud on mobile.
Factor #2. The volume of purchased traffic
Anti-fraud systems offer different payment rates, so when choosing, it is important to take into account the volume of purchased traffic to connect a favorable tariff.
Factor #3. Format of work: inhouse or with the help of partners
When working in-house, the task of anti-fraud optimization is not delegated, so there is no danger of mixing illegitimate traffic. However, this format may be more expensive or produce smaller volumes.
In cooperation with agencies, fraud filtering is one of the stages of traffic optimization. That’s why it is important to share access to the anti-fraud system. This way an agency will be able to monitor and maintain the quality of traffic.
In the case of working with partner networks, the share of low-quality traffic can increase. This happens due to cooperation with unscrupulous partners who are trying to use fraud to increase their profits. That’s why it’s important to build trust between the client and the contractor, set objective KPIs for the campaign, and use anti-fraud systems.
When working with external partners, the best option is to choose such anti-fraud systems that give partners access, so they can perform high-quality work on their side. When using anti-fraud systems only in-house, you can ignore the variety of access levels.
Factor #4. The sites where you will purchase traffic (last but not least!)
It will be enough for you to connect a system that fights the most popular types of fraud if you are going to work on channels with a low (up to 5%) share of fraud. These channels are Facebook & Instagram, Apple Search Ads, TikTok, Google. When promoting on other channels, you should pay attention to more specialized anti-fraud systems.
We have analyzed which types of fraud can be eliminated by the main anti-fraud systems:
On websites with a smaller flow of users, there is an active struggle for traffic, so you can find rarer ad fraud types there. If you work with several platforms, use anti-fraud systems that can recognize not only the most popular fraudulent actions.
For example, SpiderLabs or Admon are able to deal with Background Ad Activity, and Scalarr may be suitable for fighting Mixed Fraud.
And if you need protection from all known types of ad fraud, you can use, for example, the following combination of anti-fraud solutions:
Conclusion
Ad fraud can seriously affect the effectiveness of your ad campaigns. Use our material to minimize losses and pay only for clean traffic. Or write to us by email at hi@gomobile.ru — we will save your campaigns and advertising budgets.
