B2B E-Commerce Search Transformation: Success Stories and Training

Customer Success, Insights / AI, Search Best Practice

Background

MC+A engaged with one of the world’s largest wholesale food distribution companies to assess and accelerate improvements to their search platform. Their search powered B2B e-commerce platform services $60B dollars a year in transactions. These transactions come from two sources, internal sales associates and external business customers. The large product catalog combined with complex supplier channel sales contract restrictions presented challenges for the product team to provide a seamless and relevant search experience with compelling recommendations while maintaining contract compliance.

The Engagement

During the Search Best Practice portion of the engagement, we identified several key areas of challenge (opportunity): enhance relevancy through relevance engineering, establishment of governance over relevance engineering to monitor and report progress to stakeholders and improving the underperforming Add to Cart (ATC) events from search recommendations. These needed to be delivered while maintaining compliance with the complex supplier channel sales restrictions based on customer contract terms.

Our strategy to solve these challenges involved three major activities: Platform Optimization, Relevancy Engineering which included AI-powered search through Vectorization, and the establishment of Platform (Search) Governance.
Optimization began with our search engineers creating a defining ranking feature and the development of a golden relevance dataset from past searches, clicks, and add to cart analytic data. This led to the refactoring of a search query that was more performant (i.e. faster) and more relevant.
During Platform Optimization, we assisted with development of an operational plan to upgrade from their version of Elasticsearch to the latest and reviewed the existing ingestion process and platform configuration (mappings, etc.). This review led to configuration and process changes that resulted in an 80% reduction in index size. This optimization alone was calculated to deliver a six-figure savings in operational expenses, demonstrating the immediate value from the engagement.

After index optimization and completing a baseline of the system’s search relevancy, we focused on improving lexical search relevance with a hybrid approach to vector search. This hybrid approach was vital in managing the cost of a full vectorization of a corpus of this size. With the baseline relevance improved, we then developed a custom query intent / routing solution that efficiently served the most performant queries based on what a user was looking for. In numerous instances, users were entering product numbers, which do not work well with vector-based queries. These queries were routed to the newly tuned lexical query for additional performance gains. For more complex queries, we designed a vector search solution that significantly improved performance for most of the ‘no result’ or otherwise underperforming search queries. The experience of enhanced relevancy, increased user satisfaction and improved conversions from existing traffic as you can’t buy what you can’t find.

Throughout the engagement, governance was top of mind for the product team. Reporting, tracking and testing changes were made to provide the product team data in a manner that helped them objectively report to stakeholders the success and current trajectory of their search functionality.

Finally, to ensure the long-term success and self-sufficiency of the client’s project team, we provided comprehensive custom training led by the MC+A search architect for the engagement. This multi day training focused on enabling the relevancy engineering process, foundational search knowledge, and search best practices. The training enabled them to effectively maintain and optimize the system independently going forward.

Outcome

The impact of our engagement was substantial. The client experienced a significant increase in conversions from their existing traffic, as users quickly found the products they searched for and presented via suggestions.
Our client’s success demonstrates the impact transformation can have using best practice consulting and a smart approach to vector search for B2B e-commerce.
The results speak for themselves – increased conversions, enhanced customer experience while providing significant cost savings. As the client continues to thrive in the ever-evolving e-commerce landscape, they will continue to iterate on the immense potential of these technologies in driving future business success.

Ready to transform your business with AI?

Contact MC+A today to explore how our expertise in relevancy engineering and search technology can drive efficiency, savings, and enhanced user experiences for your business.

Customer Success Vitals

Industry

Wholesale

Use Case

B2B e-Commerce

Platform:

Elasticsearch

Are you Ready for AI?

Schedule a Discovery Call

Recent Insights

Introducing Quintus: AI-Native Investigative Intelligence

Modern Investigations Require Modern Tools From courtrooms to compliance offices to police precincts, investigative professionals are drowning in digital evidence. Documents, emails, chats, images, and recordings pile up across disconnected systems, while the demand for speed, accuracy, and defensibility only grows. Traditional review tools can’t keep up. They were built for smaller data sets and linear workflows, not the scale

Read More »

Why PostgreSQL Search Isn’t Enough: A Case for Purpose-Built Retrieval Systems 

Instacart’s recent blog posts and InfoQ coverage paint a picture of a simplified, cost-effective search architecture built entirely on PostgreSQL. It’s a clever consolidation — but also a cautionary tale. For most organizations, especially those with complex catalogs, high query diversity, or real-time ranking needs, this approach is not just suboptimal — it’s misleading. Postgres is a relational database, not

Read More »
Scroll to Top