An e-commerce site’s search results are not surfacing the items customers are most likely to buy. The company wants the simplest, lowest-operational-effort fix to present results ranked by what customers are most likely to want. Which solution meets that need with the least operational overhead?

The correct answer is to use Amazon CloudSearch and sort results by the built-in search relevance score. CloudSearch is a managed service that provides full-text search capabilities, including a configurable relevance scoring system. This approach offers the lowest operational overhead because it leverages a fully managed service with built-in relevance ranking, requiring minimal setup and maintenance. Using SageMaker BlazingText or XGBoost would involve developing and deploying custom machine learning models, which significantly increases operational complexity and effort compared to a managed search service. Sorting results by geographic location, while possible with CloudSearch, does not address the core problem of ranking by customer preference or purchase likelihood, as it's not a direct measure of relevance for most e-commerce searches.