Zvelo, a firm that specializes in detecting malicious data, is able to determine whether a visitor to a website is an actual customer looking for more information on a product, or whether the visitor is a bot(non-human) that is visiting the site for non-productive purposes, such as individuals that are paid to visit sites and view ads quickly to generate perceived traffic.
The Denver based firm says that it free bot detection service is able to inspect individual browser activity, such as how the mouse moves across the page, duration spent on a page, the percentage of the web page that is visible on the browser, the properties of the browser, among other traits. By inspecting this browser information, the company is able to determine if the user is a genuine customer, a softwarebot (non-human), or low quality traffic(LQT). This information is then put together nicely in a dashboard that is easy to understand.
If the advertiser using Zvelo’s free product finds that there is non-human traffic visiting the website, then the advertiser can pay Zvelo a fee, and Zvelo will provide the advertiser with specific IP addresses of the offending bots/LQT.
Once the IP’s of the non-desirable traffic is determined, the advertiser can then use a DSP(demand-side platform) to deny access in real time to the IP’s that have been identified. The free program will update the list of non-human IP traffic every 15 minutes, since software bots generally change their IP’s quite often.
According to Zvelo, their 3 month beta testing showed that their service has over a 90% accuracy for detection rate of bots. They found that most of their customers experience LQT traffic in the range of 5-10 %, and bot traffic of 10-40% percent.
