I've written a C# application that retrieves financial information from a public site. It'll make an HTTP request once every three seconds, twelve hours per day. That's a total of 14,400 hits per day. Using a set of ten proxies from BuyProxies.org should give me the anonymity I'm looking for (correct?). My concern is that the web site operators will learn that they're being scraped and implement protective measures. I'm taking precautions of my own to stay "under the radar" and would like your opinion on whether my precautions are reasonable or flawed. (1) My goal is to generate traffic patterns (across ten proxy IPs) that increase their site's daily traffic by no more than 1%. The assumption is that a 1% daily increase won't raise any red flags on their end. Is 1% reasonable ? (2) For my 1% figure to make any sense, I'd need to obtain a reasonable approximation of the total number of hits their site is ALREADY getting. How do I go about doing that? Here's what I've done so far: I've gone to Alexa.com and determined that each day, .01170% of the "global Internet population" views the site in question. Then, from InternetWordStats.com, I determined that the "global Internet population" is 2,405,518,376. This means the site is getting 281,446 users per day. Alexa says that each user executes an average of 6.4 page views per session? bringing the estimated total page views per day to 1,801,252. Does the logic make sense? If I'm executing 14,400 HTTP queries per day, their overall daily hit count will increase by only .80%. That's a modest increase? but modest enough not to raise a red flag? This is my first shot at screen scraping so sorry if my questions seem naive. Appreciate your thoughts. Thanks!