Understanding online users by segmenting their search keywords: empirical analysis from online auto insurance search advertising

Online search advertising is an important market considering its size and future growth potential. However, many marketers have not systematically employed the online search keywords data in developing efficient online search advertisements. While previous studies showed that search keywords can deliver intent of online users, to our knowledge, there have been few studies that empirically predict and analyze the behaviour of online users by segmenting their search keywords. Given this gap, this research contributes to online search advertising literature by examining segment-specific differences in terms of online users' behaviours and advertisers' costs. By using text analysis, we classify 9,355 search keywords from a major online auto insurance company in South Korea into four segments. Our results demonstrate that each segment shows different click behaviours at each stage of the online purchase decision process. In particular, the segment typing a specific brand name shows significantly higher click through rate (CTR) and estimate-to-purchase conversion rate (EPR) than other segments. On the other hand, the two segments typing specific product name or typing price-related words, who might be under a higher competition, show significantly higher click-to-estimate conversion rate (CER) than other segments. We also found that costs are significantly different across the four segments and advertisers pay more for the two segments under a higher competition. Thus, marketers will be able to improve their targeting strategies or advertising efficiency by redistributing their advertising budget across segments identified by the search keywords.