Fraud in mobile marketing is staggering! All marketing managers experience it daily:
- 75% of clicks are « unexplained » according to Apsalar
- 40% of clicks are “ fraudulent” base on a study by Trademob
- In 2015 $1 Billion marketing budget was wasted through bots generating fake apps installs according to Forensiq.
« We have suffered from mobile fraud for years, but we have been experiencing a sharp increase in fraudulent practices and traffic since Q2 2016», explains Hélène Queriault, VP Operations at Addict Mobile. « All ad networks and tracking tools confirm that fraud is blowing up in 2016. The fraudsters use more sophisticated technologies and it becomes more and more difficult to identify them »
What is Fraud in mobile marketing? What you should not pay as an advertiser!
Fraud is generated by shady publishers (that is to say owners of mobile apps and mobile websites), who claim to display your advertising on their app/website and to generate clicks and downloads, which are then paid by advertising networks.
There are 3 kinds of fraud:
- Fake traffic: generated by bots which simulate clicks and installs.
- Diverted traffic: this is real impressions/clicks and installs of your application, but with a lower quality than what you expected and paid for. For instance: downloads in Tier 2-3 countries (Philippines, Mongolia..) when you paid for Tier 1; incentivized downloads when you wanted quality downloads; downloads recorded several times on the same mobile…
- Fraud on your organic traffic: your natural traffic is hijacked and charged as paid traffic! E.g.: millions of fake clicks are generated by a virus (click spamming) on millions of fake ads for hundreds of applications, and when a user naturally installs an application, it is attributed to a fake click. Another example: clicks are registered when users actually want to close the ads.
How to prevent fraud and save 30% on your bill? Here are 4 expert tips
Upstream, before you start your campaigns:
- Be careful when selecting your advertising partners! You will face less fraud with qualitative sources than with incentivized traffic sources. However all advertising networks can be victims of fraudulent publishers, even Admob, GdN or Inmobi. Indeed everyone can download Google or Inmobi SDK and display advertising in their app / mobile websites with limited control. It is only a posteriori that they are identified as fraudulent publishers. Selecting the right adnetworks is only the first step: you should ideally go into details and select the right publishers in each adnetwork inventory. Addict Mobile has built up a database (DMP) to select the publishers we trust.
- Use the anti-fraud features of your tracking tool. Some tracking SDKs allow to filter fraudulent behavior. For instance, they identify fraudulent IP addresses and block all installs coming from the related IP pattern. They also identify click spamming and block fraud on your organic traffic.
After starting your campaigns:
3. Monitor everyday your main KPIs: Click-through rates (CTR), download rates and conversion rates in your application (openings, registration, purchase). To be effective, you should monitor those KPIs not only at the adnetwork level, but more at the publisher level, on a daily basis. You should suspect fraud when a KPI is different for one publisher, compared to the average. For example, incentivized traffic is suspected if the conversion rate within the application is abnormally low or if the rate install / click is abnormally high. You should also be careful with any sudden volume spike on 24-48h can be suspected.
4. Analyze in depth the campaign data as soon as unusual behaviors are identified and fraud suspected. Data come from your tracking tool and/or your agency.
Fake traffic is identified by looking at:
- IP addresses : when the same IP address (or a pattern of similar IP addresses) generated a significant volume of downloads
- Device ID : when the same device ID generated several downloads
- Mobile names (« Claire’s Iphone » etc..) : when the same name is found at a high frequency
- Types of device / brands: when the same type of device is found at a high frequency
- Time between click and install: when it is too close to zero or on the contrary, too far
Diverted traffic (e.g undesired traffic) is identified by looking at:
- Timestamp of the action (or usage of VPN) : when downloads are mostly generated in at 3am, it can reflect unwanted timezones.
- Jailbroken devices: downloads on jailbroken devices are not registered by the Appstore
- Low conversion rates: can reflect unwanted incentivized traffic.
« To analyze data in depth, identify fraud and blacklist fraudulent publishers, we use our Addict Mobile® Technology », explains Hélène Queriault, VP Operations at Addict Mobile. ”We identified 33,000 fraudulent installs for Shopmium in the UK last July 2016; it was 66% of the total campaign. Incredible! Once fraud was identified, we negotiated with all concerned adnetworks and we finally save 70% of the budget”.
Fraud in mobile marketing is now a major problem for marketing managers. To avoid scams and develop profitable acquisition campaigns, technologies that are aimed at preventing and detecting fraud (tracking SDK, DMP and robust media buying technology) are paramount. Mobile experts like Addict Mobile can help you save on your mobile marketing budget.