Gymshark adds Algolia Recommend to handle crucial Black Friday period

Founded in 2012, Gymshark is the fastest-growing fitness fashion brand in the UK. The $1.45 billion company sells apparel globally through its 15 online stores, which in 2020 saw 64 million customers visit more than 1 billion pages, resulting in an incredible $500 million in revenue. In 2018, Gymshark adopted Algolia to power its search and collection pages as it moved from its previously monolithic architecture to headless commerce using Algolia, Shopify, Contentful, React, and AWS. Find out why they made the move, and how it boosted revenue and so much more here.

A Prime Opportunity to Improve Recommendations

Gymshark’s transformation didn’t stop there, continuously innovating and improving their storefronts. In 2021, the e-commerce retailer added Algolia Recommend to the mix. The results have been remarkable, especially during the critical Black Friday period. 

The company had challenges with its previous recommendation solution. Having already had great success using Algolia for a full range of e-commerce capabilities — search, analytics, rules, personalization, dynamic reranking, and more — Ben Pusey, the Software Product Owner at Gymshark responsible for the company’s e-commerce stack, saw an opportunity to evaluate its recommendation capabilities. 

Algolia Recommend is a simple, flexible API used to build AI-powered recommendations using as little as six lines of code. It results in, as Gymshark would experience better conversion rates, increased order rate, and improved customer engagement measured, especially on returning users.

In August 2021, Gymshark initially tested Algolia Recommend in its Netherland store. During a two-week period, Gymshark tested Algolia against its previous solution to get a complete picture of how it could improve performance. Through A/B testing on related products on product detail pages (PDPs), the company saw impressive results in this first round of testing:

• A 5.5% increase in order rate for customers who clicked on a product recommended by Algolia Recommend.
• 13% higher order rate from returning customers when Algolia was recommending products.
• 10% higher “add to cart rate” for returning customers with Algolia Recommend.
• Users seeing Algolia recommendations were clicking on more products: 1.4 clicks per user compared to 1.1 with the previous solution.
• A higher conversion rate with Algolia for Gymshark’s top 10 products.

For this test, Gymshark did not test filtering options, such as showing only products of a specific color or products in stock, which would further improve the recommendation quality. Gymshark expects even better as they unlock more of the full potential of Algolia Recommend, such as the customization of recommendations. 

“This is massive in terms of users actually interacting with as many products as possible,” said Kristina Christova, Insight Analyst at Gymshark. 

Moving testing to all its markets, Gymshark saw similar results, with especially strong increases around mobile customers. Despite lower site performance due to seasonality, the retailer saw significant improvement from users interacting with the recommendations, with order rates on mobile increasing by an astounding 150 percent with the addition of Algolia Recommend. 

“We’re confident that Algolia is recommending more relevant results and we’re seeing performance increases because of that,” Christova added.

New Customer Conversions on Black Friday

Going into the vital Black Friday sales season, with Algolia Recommend operating in full force, Gymshark saw the solution’s ability to shape customer behavior. 

It was expected that existing customers would come to the site with a shopping list of items in mind and previously saved wish lists, so recommendations would have a smaller impact in the days around Black Friday, according to Pusey. 

Still, Algolia Recommend helped drive business with new site users considerably: a 150 percent increase in order rate and a 32 percent increase in “add to cart” over previous periods. 

The number of users interacting with recommendations increased during Black Friday, especially on mobile which saw recommendations use grow by 16% and, even with many users coming prepared with pre-determined lists, conversions were significantly increased among those who interacted with Algolia Recommend. 

Those great results further showcased how Algolia can support and scale during bursts in traffic, a challenge Gymshark saw in the years before adopting its new infrastructure. 

“Algolia Recommend and related products worked really well on the product pages… the impact on our Black Friday sales has been significant.”

Beyond the Numbers

In addition to the improvements in site performance from Algolia Recommend, Pusey said its use adds to the appeal of Gymshark as an employer for top tech talent. As with past Algolia solutions, he said Recommend reduces the amount of manual work required for his team. In addition, he added working with product leaders like Algolia and its innovative “shiny” technologies like it offers is a boon for recruitment. 

In the months to come, Gymshark plans to improve the performance of Agolia Recommend even more for mobile users through repositioning of recommendations on its pages. Currently, users must scroll to see recommendations, and Pusey expects a repositioning will result in even more click-throughs in the already high-performing mobile space. 

Gymshark isn’t done making the most of its Algolia Recommend deployment. Next up, it will be testing Algolia’s “Frequently Bought Together” model on its cart page to drive cross-selling, upselling and increase average order value.