Posted 12 June, 2017
Last week we met up with Senior Product & Innovation Manager from Mozoo.com, Basile Wagner. Mozoo is an Advertising Network with offices in Paris, London, Hong Kong, San Francisco and Los Angeles with over 100 employees.
We wanted to find out about his latest insights on fraud in the advertising industry. How is Mozoo using FraudShield to fight advertising fraud in their business. We challenged Basile Wagner with some questions and found some interesting insights. Read on and see for yourself for a transparent insight into daily fraud problems and how to tackle them:
My name is Basile and I work at Mozoo in our newly created Innovation Lab, as a Product and Innovation Manager. I am in charge of developing, testing and industrializing our newest products.
Through Market analysis and R&D we work with our clients and publishers to understand their problems, create solutions and, where possible, turn these solutions into commercial products.
We also guarantee that Mozoo stays at the top of the ad-tech innovation game by developing exclusive partnerships with innovative partners such as 24Metrics.
Mozoo is able to run all kind of metrics across both performance and branding campaigns. CPM, CPC, CPI, CPA, CPL, CPV.
Ad Fraud is one of the toughest challenges the industry is facing. And unfortunately, advertisers are still those who pay for the dubious practices. I don’t believe we will ever get rid of fraud entirely, no matter how advanced detection tools are.
Fraudsters will always find a backdoor.
It is absolutely necessary though for media vendors to use a 3rd party Fraud Detection Solution to guarantee maximum safety.
We need to educate ourselves more about Fraud. The more you know about it, the easier it is to spot and prevent shady behaviours.
Manually: Once we decide to run a user acquisition campaign for our clients (Our most frequent scenario at Mozoo) our first move is to require as much data sharing as possible (from both our clients and publishers):
The more data we will dynamically share in the chain, the easier it will be for a trained eye or an anti-fraud platform to identify patterns and prevent fraud to occur.
Knowing the different attribution methods is a key element. Nothing is 100% fraud-proof, but a Device Id that is based on attribution greatly reduces the risk. We know our publishers and the way they run our campaigns and their average metrics. The more transparency lies between us and our partners, the easier it will be to prevent fraud and get rid of the doubtful sources. Working with trusted partners is key in the fight against Fraud.
Finally, it is essential to monitor campaigns’ conversion rates in the most detailed way. Look at per hour, per sub sources and report any unusual Conversion Rate levels. Both high and low Conversion Rates are suspicious.
Automatically: Subscribe to a trusted 3rd party Anti-Fraud Solution. Tools like Fraudshield by 24metrics are able to identify patterns and risk level data that the human eye cannot. We can trigger alerts that will notify us and identify fraud types that no advanced monitoring can. (Proxy, VPN, IP Patterns, Bots, Emulators, Fraud Database…etc.).
We have been using it for one year now. At the time, we needed a 3rd party Anti-Fraud tool to increase our global traffic quality. After testing a few different solutions, we gave our trust to FraudShield.
We were impressed by the level of granularity the solution. It was able to provide advanced information on every fraud type we can encounter.
It makes everything easier in our daily tasks and helps us to identify and justify any quality issue we might have with our publishers.
Fraudshield is a very complete and easy to use solution. Every fraudulent conversion spotted comes with all the data we need to understand. We can troubleshoot the issue and take actions accordingly. The product team is very responsive and always adding new features to make our life easier. They care about their customers and any feedback we can provide to help improve the product is welcome.
The alert feature allows us to be notified every day if there is any suspicious activity. It saves us hours of manual monitoring and optimisations.
The flexibility of the fraud detection system allows us to apply different rules depending on the type of campaign, client and vertical (maximum CR rates and session time for CPI campaigns, cookie dropping on Mweb inventory etc…). The advanced reporting tool helps us control the quality and justify fraudulent activity to our publishers. It’s always easier to trust a 3rd party data than losing hours in an endless validation process
Finally, they provide extensive trainings about general fraud behaviour and the best ways to use the platform which is very useful for us as we constantly hire new people. We always make sure everyone benefit the same level of training and information
The increasing majority of fraud types we encounter are either IP-related (patterns or duplicate) or geo-related (VPN and proxies). Fraudshield also identifies unusual Conversion Rate levels to prevent bad quality traffic.
Thanks to Fraudshield we managed to outstandingly reduce our overall fraud rate. After the first few weeks of data we were able to detect our most fraudulent sources. Some of them averaging unexpected high fraud score, which we would never have spotted without a 3rd party tool. After a substantial data check we managed to get rid of the tip of the iceberg and significantly increase our traffic quality.
IP whitelisting and callback security token in our S2S configurations is the very first step we take when we run a campaign for our clients.
We always enforce our campaigns geo targeting to filter out every non-whitelisted geo and avoid mistargeted conversions.
Then, as the campaign is running we proceed to monitor Fraud on a daily basis via the FraudShield dashboard. We make sure everything is running smoothly and only clean conversions are coming through.
Finally, when it comes to extremely sensitive campaigns we set every incoming conversion as pending. We proceed to a thorough validation before approval.
The fight against fraud must be taken extremely seriously by every actor in the value chain and everyone must take responsibility (brand, agency, ad networks, publishers, attribution solutions etc.) and at every step of the delivery process (app tagging, campaign negotiation, set up, delivery, reporting etc.). As the industry tends to require more transparency it will be more challenging for fraudsters to operate freely. Shady networks will naturally disappear.
Basile Wagner – Sr Product & Innovation Manager @MOZOO