SVG Releases Proprietary Bidding Engine

August 23, 2016 /



One commonality amongst SVG’s various departments is a shared passion for innovation. In our journey to build an industry-leading insurance distribution platform, we regularly create tools that are custom-tailored to our marketing and sales processes and help us automate and optimize tasks across multiple departments.

The most recent example of such a tool is a proprietary bidding algorithm written to maximize marketing efficiency. The program checks daily for changes in keyword behavior and automatically changes bids and other campaign settings based upon several layers of machine learning properties. This means the program becomes smarter and more accurate with its predictions over time.

A unique part of these learnings includes the ability for the algorithm to identify search terms that did not previously exist in the program but that are being used at a high rate in web form conversions that lead to policies in the sales process. As with all machine-based adjustments, pricing for these newly-identified keywords is based on a variety of past keyword judgments.

The resulting bidding is much smarter and helps us identify interested consumers more efficiently. With this program enabled, we are casting a wide net shallowly as opposed to casting a narrow net deeply. While a narrow deep net will drive high amounts of traffic, the resulting traffic likely includes a good amount of people who are less likely to become policyholders.

As our consumers become smarter and more aware of their ability to purchase insurance online, we must become smarter and more efficient with our marketing efforts. It is our responsibility as online marketers to deliver the most applicable experience to the right person at the right time.

© 2018 Spring Venture Group / Privacy Policy