ProxyGuard.com
Title
- PROXYGUARD ::
Description
Internet businesses in the advertising, search-engine, affiliate-program, or traffic-trading industries are based on the potential productivity of the visiting IP traffic (i.e. shoppers, searchers). Transactions amongst those businesses, their customers, and their partners are all premised on the ability to track the quality of the underyling IP traffic. Early tracking techniques used client-side mechanisms (i.e. JavaScript, cookies, hidden images, counters) to ensure the identity of real-life browsers. However, those techniques were quickly spoofed by saavy hackers with automated software developed to falsify this information. Likewise, monitoring techniques that analyze the "realness" of the browsing traffic behavior (i.e. expected throughput ratios, download and linking to content in a realistic amount of time) are spoofable by software robots designed to emulate human behavior. Thus, the current de-facto standard used to identify the validity of IP traffic relies on the uniqueness of the requesting IP address and monitors the frequency of unique IP usage with a particular time window. This standard, however, is susceptible to fraud through through the use of innumerable proxy servers distributed around the world.
Proxy servers, when correctly used and implemented, are a powerful and necessary tool to improve networking, routing, security, and information caching on the internet. Many corporations, universities and ISPs use proxy servers for legitimate purposes. However, the abuse of misconfigured servers allows automated robots to mask their own IP with that of the proxy server, thereby giving a single computer the identity of thousands of seemingly unique and independent people. In addition, the identity and number of proxy servers is constantly changing: new proxies come online and expire all the time. Coupled with the real-life behavior-emulation and falsification techniques mentioned above, modern software agents use these dynamic proxy lists to become, for all intents and purposes, nearly indistinguishable from their human counterparts.
Any internet business paying for unique traffic, be that cost-per-click, cost-per-search, or cost-per-ad--, is at high risk. In our experience, fraud levels for cash-payout programs can often exceed 40%. Saavy cheaters also tend to "spread the fraud around". Rather than defrauding a single program for a high daily amount and immediately appearing suspect, proxy robots target hundreds of separate CPC, CPS, or CPA programs for small amounts. Their cumulative fraud totals are large, but by keeping individual program payouts low, they can "stay under the radar".
