Sinequa Augments Companies with Release of New Generative AI Assistants. Learn more

Chat with Sinequa Assistant
Sinequa GenAI AssistantSinequa GenAI Assistant

What kind of search engine is best adapted to an e-commerce site?

Posted by Laurent Fanichet

What kind of search engine is best adapted to an e-commerce site?

All the experts will tell you: one of the essential services of an e-commerce site is the search tool that it proposes to visitors. This means that e-merchants have to choose the search engine to be integrated not only with their Website but also their back-office application.

The choice is far from simple since it needs to take into account three potentially contradictory objectives: offering the best possible answers to visitors’ queries; taking into account the commercial strategies of the merchant; fulfilling the requirements of the IT department in terms of performance, openness, and reliability. Facing these requirements, Business Search solutions turn out to be the most adequate for e-commerce Websites.

I have been communicating for some time with big players of this industry, like Leroy Merlin, RueDuCommerce or Pixmania. The latter is at the center of recent news since industry giant Carrefour delegated its non-food e-commerce to Pixmania. This is a point in case of our argument for Business Search, since Pixmania has been using Business Search solutions from Sinequa for a number of years.

The first advantage of a Business Search solution is to offer responses to all requests formulated by visitors of an e-commerce site. Such a feat cannot be accomplished by asking simple (keyword-based) questions to a classic search engine or formulating SQL queries for a relational database.

Indeed, the latter case would require the internet customer to formulate queries respecting the terminology or precise product names of the vendor. If the internet customer makes a mistake, the simple search engine or the database will come back with an error message – creating frustration for the potential customers and incite them to leave the site and look elsewhere.

A Business Search solution adapts natively to the commercial strategies of the e-merchants

Due to the influence of Google, Internet users want to ask questions without constraints. A Business Search solution is able to interpret requests in natural language, and to come back with not just one answer but a series of answers within the context of a user’s request, making use of all the heterogeneous information available at the e-merchant. This capability allows to not only give relevant answers to a potential buyer’s questions but also to answer imprecise questions, and to present answers ordered by context-oriented facets.

Such a solution also favours cross-selling and up-selling: by indexing the entirety of commercial information of the vendor, (list of products, stocks, promotions, etc.) it is capable of showing similar complementary products related to a user’s request.

A request for an iPhone will thus not only return the available models but also information on accessories, applications, etc. The vendor need not establish “hard-coded” a-priori links between these different offerings. A Business Search solution does that on its own, and organizes the responses intelligently, by category, by price, etc.

It is also possible to attach different “weights” to categories, thus influencing the relevance naturally allocated by the engine. This allows, for example, giving priority to promoted articles in response lists.

The e-merchant has thus complete freedom to implement special commercial strategies and ad-hoc tactics. They will be automatically taken up by the search engine. Thus, Business Search imposes itself as a solution that corresponds to the needs of customers of an e-commerce site, enhancing their customer experience, while also respecting the business imperatives of the e-merchant.

An “e-commerce search engine” must be able to process 1000 requests per second.

Finally, e-merchants also need to cope with the seasonal constraints of their business. During Christmas season, the major players in e-commerce may receive tens of millions of requests per day. Facing such challenges, only specialised search companies are able to offer sufficiently scalable solutions, supporting more than a thousand requests per second while preserving an optimum level of performance and stability.

Robustness and scalability, openness, support for the commercial strategies of the e-merchant, and customer satisfaction: here is a set of requirements for a search engine well adapted for an e-commerce site.

The topical Carrefour-Pixmania deal adds another one: A Business Search solution must adapt naturally to the B2B models and expansion strategies of e-commerce players (market places, generic brands, outsourcing), with a large number of catalogues from different vendors. This is where the capabilities of Sinequa Business Search in connecting to and indexing large volumes of heterogeneous data make a real difference.

Register to the newsletter
Sign up to our newsletter