Rapha Cycles Away from Monolith to MACH

Rapha—one of the world’s best-loved cycling apparel brands—was on a mission to perfect its digital experience. Now, as part of a move to a full MACH (microservices based, API-first, cloud-native SaaS and headless) strategy, the company has radically improved the way it works with and delivers visual media. This move to MACH reduces the time that creative and merchandising teams spend working on product and customer images and videos, optimizes page load, transformation times, and even makes it possible for them to utilize new lightweight file formats.

ABOUT

Superb images and video inspire riding without limits

Driven by a mission to inspire and equip people to live a life by bike, all day, every day, London headquartered Rapha is the world’s leader in cycling apparel. Founded by Simon Mottram in London in 2004, Rapha’s products have redefined comfort, performance, and style for cyclists from absolute beginners to WorldTour professionals. With innovative products designed for all types of terrain and rigorously tested around the world, this truly global brand aims to improve the riding experience for everyone, which it does through a mix of online and its famous ‘Clubhouse’ real-world cycling community hubs.

THE CHALLENGE

Time to upgrade an outdated system

According to Rapha Principal Engineer John Kilpatrick, the brand had simply outgrown what had once been the right fit for its e-commerce and content management needs, and the process for shipping new algorithms to handle the increased load had become too rigid.

And when most of your commerce moves online, that rigidity slows you down – “In terms of sales, digital’s our biggest channel, so our imagery and video is vital for how we bring our brands and product to life for customers,” Kilpatrick explained. A particular bottleneck had been getting assets out. Kilpatrick continued, “We saw an opportunity to speed up the ingest process to allow our merchandising and brand teams to deliver value much faster. Speed is incredibly important, especially for merchandising, which has to work ahead to get content.”

That’s why Kilpatrick and his team embarked on a strategy to build a completely new, best-of-breed, open enterprise technology ecosystem to deliver a seamless, composable architecture. Rapha’s new AWS-hosted infrastructure is a new MACH stack that includes Commerce Layer’s API-rich commerce platform and Contentful’s headless CMS. Cloudinary sits at the heart of the tech stack, managing all of the brand’s asset management and delivery processes.

For now, this new composable architecture operates in parallel with Rapha’s legacy stack, but Rapha aims to move more and more business processes to MACH structures over time.

Right from the start, Rapha envisioned a modern DAM that would support a transition to MACH. Cloudinary came up very early in the conversation, says Kilpatrick, and was selected over three rivals after a very thorough proof of concept. Cloudinary is now the main image handling workflow for Rapha’s growing library of 25,000 assets, of which around 70% is images and the rest video, including some great rider user generated content (UGC). What sealed the deal with Cloudinary is that it promised to speed up image work out of the box and provided built-in transformations.

THE RESULTS

Page loads and image wrangling take a flyer

Cloudinary has transformed the way the company works with images from go-live. Crucially, getting media assets out is instantaneous.


But even better, the Rapha team has been able to speed image improvement and delivery work by at least 90% for images and 50% for video, he confirms. Acceleration is enhanced by being able to use a whole range of lightweight image formats instead of heavyweight JPG files alone. These lightweight formats consume up to 95% less digital space. Cloudinary has also driven huge improvements in image rendering times, which boost SEO scores and keep customers engaged:


• ‘Visually complete’ metric improved by 50%
• Hero rendering times are reduced by 70%
• Largest Contentful Paint (a Google Core Web Vital metric) is 20% faster
• Page load times run 40% faster
• Image loads are 70-90% faster

For a website that is so image-driven above the fold, improvements in image rendering times have immediately and dramatically decreased page load times. These enhancements are not only a huge boost for the brand’s highly important Core Web Vitals score, they have revolutionized the way Rapha works with images across the company. Thanks to Cloudinary’s DAM, common processes that took up to six hours now take three.

“The team managing assets is spending half the time because of the mix of the presets and all of the automated tagging, collections, and sophisticated metadata they get with Cloudinary. And as the URL names make sense too, all their search and workflow is so much better, too,” Kilpatrick states.

Image optimization is a key win, too. The Rapha team is a big fan of the Cloudinary automatic format selection (f_auto) transformation, which automatically delivers images in the most optimized format.

The constant tech guidance Kilpatrick and his team has received from Cloudinary along the way, especially around EDA transformation, created a slipstream that enabled Rapha to glide through implementation. When tallying up all these benefits and more, like a better way to store images, it’s clear that Cloudinary, as part of a composable architecture, has won the coveted yellow jersey.