The Black Friday that Broke Gymshark’s Store

Gymshark’s urgent need to adopt headless commerce came after Black Friday 2015. The surge in customers during this prime sales event overwhelmed its existing self-hosted store, resulting in a crash that not only cost Gymshark revenue but also a hard-won reputation. 

A decision was made to move to a headless commerce architecture, prioritizing best-in-class, scalable, API-first technologies. From its monolithic store, Gymshark selected Shopify as the foundation on which it built a new storefront. It then adopted leading solutions for search and navigation capabilities (Algolia), front-end interface (ReactJS), and customer relationship management (Contentful), all supported with additional AWS services and tools.

A Serious Need to Improve Product Discovery

Peak customer loads weren’t the only driver for Gymshark’s transformation. The store’s search and discovery functionality had “big problems,” according to Ben Pusey, Software Product Owner at Gymshark responsible for the company’s e-commerce stack. 

Poor search results meant best sellers were being buried at the bottom of the page, while out-of-stock products would appear at the top. Low relevance led to poor conversions from search. And the shopping experience lacked any of the personalization that consumers have come to expect. In addition, manual merchandising was intensive and unscalable for the merchandising team. A lack of localization resulted in slow load times for shoppers outside the U.K. and poor search results based on language nuances.

The Solution: Improving Relevance through Algolia

The company turned to Algolia to improve site search, browsing capabilities on its product category pages, to reduce the heavy workflow demands on merchandising, and more. Pusey says Gymshark chose Algolia based on four factors: 

• Ease of development: Well-documented and easy-to-use APIs.
• Scalability: Ability to handle peak traffic during sales events.
• Innovation: Continuous introduction of new features and capabilities.
• Collaboration: Working closely through beta tests and the development of new features. 

Implementation was fast: Gymshark moved from its monolithic self-hosted store to headless in a matter of months rather than years, and it deployed in-house personalization using Algolia in just a quarter. The company is now using Algolia’s entire relevance stack to meet several business goals and looking at how it can take further advantage of it.

Boosting Relevance for a Superb Search Experience

Today Gymshark uses Algolia for site search to unlock incremental revenue by customizing the user experience to their specific needs, audience, and catalog. But its first priority was to improve relevance. Gymshark uses Algolia for site search to unlock incremental revenue by tailoring the customer experience to their specific needs, audience, and catalog. 

The team added criteria to their ranking such as product availability, preventing out-of-stock items from showing up first on search pages. They used other business data to fine-tune relevance on a query basis, for example, improving the average click position of products based on its most search keyword “camo”, which drove higher click-through-rates. In addition, Gymshark used Algolia’s AI-generated synonyms to show customers relevant products previously being missed. For instance, U.S. customers searching for “sweatpants” were getting no results, since in the U.K. they were called “joggers.” Algolia’s AI detected — and corrected — this nuance. 

Algolia’s AI-based merchandising capabilities were used to promote best-selling products based on each individual search query, which has resulted in an estimated extra £2 million per year in sales. 

The overall results are remarkable. In just a year, search conversion rates went from 6.2 percent to over 10 percent and climbing. And the improved experience that Search now brings has more customers using it. While search once played a part in less than 10 percent of orders, on Black Friday 2020 it was used in more than 30 percent. Revenue from search users is up more than 400 percent year over year!

Increasing Customer Engagement with Algolia Powering Collection Page

The team saw such an impact from Algolia on its site search, it looked to apply its benefits to the customer navigation experience. It now powers its product listing pages (PLPs) and collection pages with the Algolia relevance stack, bringing the benefits of custom ranking, Rules, Dynamic Re-Ranking, and the ability to combine human and AI-led Merchandising capabilities. 

Gymshark’s merchandising team now uses Algolia’s Visual Editor to define precise merchandising rules based on the company’s business data, automatically merchandising across its entire store — improving the customer experience through speed and relevance. 

Best sellers and items with the most sizes in stock are prioritized, while out-of-stock items are hidden using a priority scoring system for product ranking. Pusey estimates the company has generated £4.5m a year in extra sales. 

In 2021, the company tested the effect of Algolia’s relevance capabilities on its leggings product page. The category is its highest performer and has an incredibly large range of products to support (nearly 60 legging products and 223 color variations.) 

The results were remarkable. By promoting well-established items at the top of the page, those items became responsible for 8.3 percent of click-throughs and 15 percent of overall page revenue. Meanwhile, without it, the top listing was responsible for a mere 3.2 percent of clicks and 3.9 percent of revenue. 

As well, orders were abandoned approximately 10 times less often and items were added at checkout about 40 percent more frequently. Revenue per click increased by 8 percent, but even more importantly, purchases made were on items that saw less returns. Stats show fewer returns translate directly into better customer loyalty.

Reducing manual workloads with AI Merchandising

While Algolia solutions improved customer experience and improved sales, adding AI merchandising capabilities provided a boon for Gymshark internally. By automating merchandising, Algolia’s visual merchandising tool reduced manual labor, but also helped Gymshark overcome reliability issues Pusey says were causing system crashes. 

“During really really busy periods the trading team wasn’t able to merchandize the site fast enough and react to things coming in and out of stock, and what was being bought and what the trends were,” Pusey noted. 

Everything is now rules-based, allowing changes to be made in near real-time to account for things like stock levels, new products, and consumer trends. Merchandising has gone from a heavily manual process to “basically set and forget.” 

It has allowed its trading team to avoid scrambles and focus attention on other activities to provide value, and scalable and reliable technology prevents the crashes — and resulting customer loss — it once experienced.

Personalization and Beyond

Gymshark is using Algolia to personalize the shopping experience to gain incremental revenue and build brand loyalty. Initially implementing it relatively simply across search and collection pages the company expanded use of Algolia’s personalization capabilities to across the entire website and even outbound communications. 

Gymshark started by creating personalized search and merchandising placement using customers’ color preferences, using the information from their shopping carts, but plans to incorporate personalization around specific events, pricing tiers and more — using deeper data and analytics. 

It is now exploring how it can use personalization to reap even great benefits through driving personalized recommendations and cart upsells.

Increased Revenue is Only the Start

Pusey estimates that adopting Algolia for search, improved customer navigation and AI-based merchandising have all resulted in an astounding $20M incremental annual revenue. 

While transforming its e-commerce architecture by adopting a headless commerce approach and Algolia for search, navigation, AI-led merchandising and personalization has allowed Gymshark to grow revenue, the impact goes far beyond red and black numbers. 

By taking a headless and microservice approach, Gymshark is ready to evolve its e-commerce capabilities incrementally through an agile approach. For search and navigation, its next steps are to test KPI-driven merchandising algorithms, apply machine learning reranking to collection pages, test new personalization strategies, and implement product recommendations, all while launching a mobile app benefiting from all those capabilities. For the rest of the stack, Gymshark is gearing up to implement a new product information management system. 

It’s already experienced tremendous success, and is looking to how it can use Algolia and other top-tier solutions to improve the customer experience in the days, weeks and years ahead.