Nosto’s Search engine finds search results by matching user queries to the product data defined in Searchable Fields (Search → Settings). As soon as the product data is initially indexed in the Nosto Product Intelligence and the Search module is enabled, you can start setting up Searchable Fields and their priorities. Your onboarding manager will help you to find the best solution based on our experience and your individual business logic.
Best Practice:
Basic searchable fields: product name/title, category/product type, brand, important attributes (e.g. color)
Searchable fields should only contain data fields with concise information (no long product descriptions, SEO descriptions, etc.)
Define high priority for max. 5 different fields to be able to emphasize their importance
Don't use more than 10 different searchable fields as this might lead to a lot of unspecific results due to false matches
Nosto’s search engine constantly processes product data in a backend by optimizing, cleaning, and ensuring search relevance.
On query time, Nosto’s search engine uses various Natural Language Processing (NLP) algorithms to understand the user’s intent. Examples would be stemming algorithms to understand that words with the same stem have the same meaning (e.g., knife - knives, deportivos - de deporte), compound word algorithms, (e.g., damenjacken - jacken für damen) or the automatic detection of singular/plural forms in any of the supported languages out of the box.
On top of the keyword matches, the search engine applies matching merchandising rules based on attributes, performance metrics and various personalization and affinity metrics to reflect users behviour and ensure that search results are relevant to the current shopper and ranked by business needs.
Learn more about search, merchandising and personalisation here.
Learn more about how to use score insights to understand how to analyse and optimise the sorting of product listings here.