Already today we’ve seen a handful of responses to yesterday’s story about fraud detection tools. So, in hopes of bringing a number of various perspectives to light and create further, more insightful dialogue on the topic, we published a post today featuring comments from Smarter Chaos’ Matt Frary and CPAWAY’s Thomas Dietzel. Now, we’ve got a follow-up reply from CPA Detective’s David Sendroff supporting his product and refuting the earlier claims made by Dietzel.
In response to Dietzel, Sendroff stated,
“Most networks value long-term relationships with their advertisers and proactively use tools like CPA Detective to identify fraudulent traffic sources. In the long run, eliminating bad traffic will maximize profitability for all legitimate players involved—advertisers, networks, and publishers.
The value of a third party tool like CPA Detective is gaining access to a centralized fraud database based on global patterns across the affiliate ecosystem. Our proprietary methods of detecting fraud and malicious proxies, without leveraging publicly available services, allow us to truly understand why each element is flagged while maintaining a low false positive rate.
Widely available proxy lookup services are no longer an effective measurement of traffic quality because fraudsters are commonly checking IP reputation to see if a proxy is listed before submitting stolen or recycled user information for their affiliate commissions.
While we provide transparency and the ability to export granular details of each conversion that is flagged, clients choose when and how to present the results based on their own policies and risk tolerance. We do make clients aware that the more that you show, the more that fraudsters will attempt to exploit the system, so they can use their discretion.
We are looking to make this a more sustainable channel. Tools like ours allow advertisers to minimize risk, gain confidence in the power and ROI of affiliate marketing and allocate their budget on programs that they may otherwise avoid.”
In the end, relationships with advertisers are key, and these types of tools can be very useful in cleaning out fraud. However, the ways in which traffic signals are interpreted by the tools and how those results are then interpreted by the users remains a topic for debate.
What do you think about these fraud detection tools? Let us know in the comments.