One important strategy few online retailers tap points to tying Web search with Web site query data and using cloud services to support targeting. A study from Microsoft Research, "Data Services for E-tailers Leveraging Web Search Engine Assets," analyzes the options and presents three cloud data services--entity synonym, query to entity, and entity tagging -- to support smaller etailers whose marketing execs don't believe their company can match the same rich user experiences as larger online retailers.
The entity synonym service is available today through Microsoft, while the other two are still in development.
As Microsoft Research analysts point out, online retailers like Amazon have many consumers visiting and shopping at their Web sites. Activities yield tons of data related to queries, search results, and products bought. While large retailers mine the data, smaller counterparts don't. Maybe the process proves too expensive. Whatever the reason, Microsoft believes several cloud services it has begun to develop can bridge the gap.
The paper steps through each of the data services. The Query-to-Entity service, for example, requires the ability to tap into the query click log from Bing between 2010 and 2012. This data service even evaluates the latency of online auto-completion based on the queries discovered by the entities. "To support online auto completion, we implemented an in-memory Trie with Windows Azure as the backend infrastructure. Our results show that we can auto-complete 99.9% of the queries under 100ms for a query log consisting of over 100 million distinct queries. This is a 'worst case scenario' since most e-tailers' query logs have far less queries," the paper explains.
The paper aims to make the connection between search engine data assets and mining information for e-tailer's products. Microsoft analysts also present the idea in the paper of using search engine data assets to enhance the user experience for product search on retail Web sites. The three cloud data services aim to enhance the user experience for product search on their sites.