Search Engine Marketing (SEM) can help enterprises show brand, promote activities, and drainage to their own platform, bring business transformation, so enterprises pay more attention to the investment of SEM, and put forward higher requirements for effect evaluation, especially value the direct transformation brought by SEM. But if only the value of SEM is equal to the direct conversion rate of this single index, it will “see the leaves, not the forest”, seriously underestimate the real value of SEM, and into the SEM optimization is difficult to have a new breakthrough, and even affect the overall strategy and layout of enterprise marketing. When serving an insurance enterprise, facing the above industry limitations, based on the advantages of advanced technology and big data, HORi Network helped the enterprise effectively restore the real value of SEM through the analysis method of “large attribution + small attribution”, and put forward a new idea for the optimization of SEM.
Big attribution: reducing the true role of SEM source + bridge + direct transformation
In view of customers’ doubts about the value of SEM, HORi Network used SQL language to retrieve the marketing data in OpenBI, an open data analysis and mining platform set up by Hori Network on GDMP, and conducted attribution analysis on this basis to clearly restore the real value of SEM in a visual way.
On the one hand, we used OpenBI platform to collect data from the first arrival crowd, final conversion time and order quantity of an insurance customer’s official website to analyze the decision-making cycle of the insurance industry crowd. The analysis found that the average conversion period of the insurance industry is 7 days, 43% of people do not buy on the same day, so the real effect of SEM should be measured in a relatively long period of time.
On the other hand, we collected bridge words and source words from paid search, non-paid search and direct source channels, and transformed data for further attribution analysis. SEM traffic directly contributes to the transformation only accounts for 25%, and in the other 75% of indirect transformation traffic, SEM channel as a bridge and source traffic accounts for 45%. If we only focus on direct transformation, we will ignore nearly half of indirect transformation contribution, and indirectly affect the implementation and adjustment of subsequent marketing strategy of enterprises.
Small attribution: dig the keyword value deeply and optimize SEM to the extreme
After restoring the real value of SEM channels, we made further optimization and expansion on the conventional SEM ideas, that is, attributing according to the particle size of SEM keywords, analyzing the assist effect of non-transformed keywords, and innovatively proposing optimization indexes such as assist rate and single point rate to effectively quantify the assist effect.
In the first stage, the key words’ jump paths are connected into a graph model, and the granularity of the key words is analyzed according to the graph attribution technology of jump probability calculation of transformation possibility, and the promotion effect of the key words on transformation is quantified by the innovatively proposed index of assist rate and single point rate. The higher the assist rate, the greater the promotion effect of keywords on the transformation effect, while the lower the assist rate and the higher the single point rate, the less the promotion effect on the transformation effect. In the second stage, according to the graph attribution results of the first stage, the granularity of time-sharing keywords is further analyzed, and it is suggested that customers only open the period of high clicks and high assist rate, and suspend the delivery in the rest period.
In the one-month first-stage effect test, the budget of keywords with an assist rate of more than 10% in the history of three months was increased, and the keywords with a single point rate of more than 95% and an assist rate of less than 3% were suspended. Compared with before the test, the average daily consumption and average daily order quantity increased after the test, while the order cost decreased significantly. After the second stage of optimization, the test effect is also very significant, further improve the utilization rate of the overall budget.
Through the attributive method of OpenBI+SQL language, I first analyzed the life cycle of different types of users through the network, developed marketing strategies in line with the segmentation group, and restored the real value of SEM. And by quantifying the assist function of different types of keywords, the optimization and innovation of SEM are realized. In addition, the method can also be applied to help enterprises analyze the long-term value contribution of short-term activities, the advertising overlap degree of various media groups, and guide optimization strategies through the analysis of keywords, advertising creativity, advertising style in a more fine-grained way. Based on the unique distributed data architecture of the HBORI Big Data platform and advanced real-time, multi-dimensional correlation analysis technology, Hbori’s solutions enable customers to fully understand the complex relationships between data, gain new business insights, and make better business decisions.