How LuckyVitamin Increased Revenue by 13% with Constructor

“Constructor empowers us to curate a merchandising experience. We always wanted to, but we could never move past step one before.”

LuckyVitamin is one of the world’s leading health and wellness retailers. A family-owned, brick-and-mortar pharmacy that launched its first website in 2004, they’ve since skyrocketed—now offering more than 13,000 products to more than 50 countries across the globe.

The brand is committed to helping customers and site visitors alike live happier and healthier lives through a range of products, including vitamins and supplements, personal care and beauty, and grocery items.

Lucky Vitamin

The Problem

LuckyVitamin had previously relied on another search vendor to handle its site search. While search would work for customers who were already well-informed as to the exact product they needed, the vendor’s string matching algorithm wasn’t prepared for customers who would enter more general queries, for example around “heart health” or “keto diet,” which was a common occurrence (and became even more common when COVID-19 hit). 

It wasn’t uncommon for search results to provide either hundreds of results or none, and customers began to get frustrated and abandon the site without finding what they were looking for. 

“Our customers would get something like 210 results, and that experience left them wondering what to do next. They had no way of comprehending which product was the right choice,” explains Vishal Arora, Senior Director of Digital Experience at LuckyVitamin.

Vishal Arora

Director, Digital Product & User Experience at LuckyVitamin

Connecting Customers to Products through AI

With customers abandoning and losing trust in search, it was time to seek out a new solution. Search and browse were critical functions to LuckyVitamin’s customer experience and revenue, and drops in performance on those pages threatened to impact both the customer experience and the bottom line. 

After broadly evaluating a number of product discovery options, the LuckyVitamin team found that Constructor’s approach to search and discovery was fundamentally aligned with their vision. Together, Constructor and LuckyVitamin were able to complete an end-to-end integration in just a month. Looking back on the process, Arora reflects that implementation was “an incredibly strong, incredibly important, and validating experience.”

Tailoring search results and browse pages for each site visitor based on that user’s clickstream data, Constructor’s machine-learning platform helped surface products not only on textual relevance but on attractiveness as well. 

“We used to miss so many connections between conversational search terms and the products they were intended to describe,” Arora says. “With Constructor, we can actually get those items in front of people because the connection between actual behavior and search terms is stronger.”


Harness clickstream data to show results that are both relevant and attractive to individual users, increasing conversions and lifting business KPIs.

Group 745 (1)

Get your customers where they want to go faster. Autosuggest infers search intent faster than the user can type their query, so you can capture attention and increase conversion rates.


AI-powered, personalized category pages and smart merchant tooling serve products that make users happy and contribute to your bottom line.

The Results


lift in revenue


lift in search usage


lift in conversion rate

Almost immediately, Arora began to see dramatic lifts in key metrics with Constructor’s AI and machine learning implementation for search, autosuggest, and browse.

“The performance obviously speaks for itself,” he says. “We were expecting a lift, but not to the degree that we’ve seen it.”

And with Constructor’s machine-learning and AI-backed approach to product discovery, the lifts keep growing.

Fighting Zero Results, One Search Query at a Time

The world of holistic wellness can be complex from a search perspective because vitamins and supplements often have different names for the same product (such as curcumin and turmeric) and multiple uses for multiple regimens. Building in all of these considerations became a highly manual and frustrating process for LuckyVitamin’s merchandising team.


Customers would often search for a generic product but get either wrong results or none, which led to a lot of bounces from the site—and lots of potential lost revenue. While the team would mitigate with synonyms and redirects, it soon became a problem of scale in a catalog of more than 13,000 SKUs. Customer confidence in LuckyVitamin’s search diminished over time as customers began to prefer browsing category pages for items rather than searching for them. 

This is exactly the type of problem that Constructor’s Cognitive Embeddings Search was designed to solve. Using natural language processing to expand search coverage and prevent zero-results searches, LuckyVitamin was able to steer customers to the right products, significantly improve their conversion rate, and rebuild customer confidence in search.

Lucky Vitamin Keto Diet Screen

“Going back to our zero-result reports from before and trying some of those same searches today, it makes complete sense why people are converting more because of it,” explains Arora. “Instead of giving them weird results or no results for when they search for something that’s reasonable, we steer them away from that dead end. Customers then explore, go into the category pages, build bigger baskets, and actually convert.”

The LuckyVitamin team also found that with Constructor’s AI and machine-learning platform offering data-driven insights and automating much of the hands-on merchandising work, they were able to be more forward-thinking and strategic in ways to provide value to their customers.

“Constructor lets our merchandisers do merchandising strategy. Instead of trying to be dictionaries, we can now focus on the things that we’re experts in. Our work today involves talking about relative rankings, like ‘what should the merchandising rules be for this term’ and not questions like whether an item should even be on the page,” Arora says.

Manage by Exception with Merchant Tools

Constructor’s searchandising platform allows merchandisers to set global, category, or attribute-based rules for easy management that does away with a lot of manual work. But for specific situations, robust merchant tooling allows ecommerce teams to “manage by exception” by boosting, burying, and slotting items as needed.

What’s Next for LuckyVitamin

When it comes to LuckyVitamin’s goals and roadmap for the future, Arora knows where he wants to go. “We want to be an educational resource as much as an ecommerce site,” he says. “People don’t just buy supplements on a whim. They want to make sure that they’re safe, they’re made with good-quality ingredients, and they’re going to meet their health goals. We take that pretty seriously, and we want to be able to give people that empowerment to make those decisions for themselves.”

A huge part of this initiative involves leaning into search. With search and browse now being a reliable and essential contributor to LuckyVitamin’s business KPIs and customer experience, Arora sees an opportunity to use search results to trigger not just product results, but helpful educational content created by on-staff nutritionists and wellness experts. 

“Now that our search results are better, we can offer more guidance to our customers and be a guide to them,” he explains. “And that has broad implications for our fundamental business.”

Product Discovery As It Should Be:
Designed for Ecommerce. Driven by AI.

With every action they take on your website, your customers are telling you what they want to buy. Constructor actually listens to them. 


Applying machine learning and AI to your customer clickstream data, Constructor creates personalized, optimized product discovery experiences across search, browse, recommendations, and more. Our products help top ecommerce companies drive KPIs and grow lifetime value. 


Let us quantify the value of Constructor’s ML-backed search and discovery on your site using your data.

See for Yourself

Validate Constructor on your site, using your user data, and see exactly how our platform can lift critical business KPIs.