Kurt Geiger sizes up MuleSoft APIs to deliver a perfect fit for partners
- Shoe designer can get partners up and running in just 12 weeks rather than eight months.
The retail industry has been experiencing a huge amount of turmoil over the last few years, with high street stores closing in their thousands and consumers cutting back on spending.
British shoe designer Kurt Geiger, which operates over 70 stores and 530 concessions worldwide, was already in the process of re-defining where and how it distributed goods as the pandemic hit.
This included re-evaluating its concession and wholesale partners, particularly in the UK, depending on the direction of their business strategies. Gareth Rees-John, Chief Digital Officer at Kurt Geiger, says:
There's been a huge amount of downward pressure on the retail industry over the last few years. Some of our partners had responded in ways that perhaps were not fully aligned to our brand strategy. We very much didn't want to get into a pricing war, a Black Friday frenzy with different partners and different code usage that was going on.
Pre-pandemic, Kurt Geiger had been distributing goods via partners in two main ways. The first was wholesale, where the retailer buys the stock up front and commits to a sell through. This is a well-tested method for the brand, particularly for new markets and dealing with large-scale partners such as US partners Nordstrom and Dillard, where it’s still in place to great effect.
Then there was the concession model, where the third-party distributor would hold the stock, but Kurt Geiger would retain ownership of it. This latter model brought challenges, as Rees-John explains:
One of the problems was we had all these pots of stock in different places. If you are distributing through multiple partners, you'd have stock in multiple places all upon your own ownership still. When you didn't really know exactly what was going to sell where the fastest, we came across all sorts of availability issues and sell-through problems.
The shoe maker decided to shift to dropshipping. With the dropship model, instead of a third-party partner like Farfetch or John Lewis dispatching the stock from their own warehouse to their customer, Kurt Geiger distributes products directly from its own warehouse to the partner’s customer. Or as Rees-John puts it:
They retain the customer relationship. We just do the shipping of our products to them directly.
If the shoe fits
Kurt Geiger needed technology that would enable dropshipping, letting the business plug in its stock from a single warehouse or multiple warehouses, and then distribute it across a variety of different platforms. The objective was to gain better control over the stock and more efficiency in moving goods.
One of the firm’s biggest costs is its stock inventory and therefore maximizing sell-through rates, particularly at full price, is fundamental to success. Rees-John explains:
From an EBITDA point of view, it clearly makes sense, and also from a brand values point of view - first price, right price - and getting the value/cost equation right.
The company originally looked at whether it could shift to dropshipping using its current technology stack, but that would prove tricky as it had multiple integration points and different pieces of information that would need to be linked up from its own business to its partner network.
Kurt Geiger was already a fair way along its digital transformation across the organization at this point, and had gone down the path of micro-services as it built a fully composable architecture for its current direct-to-consumer (D2C) sites. Rees-John recalls:
We were moving - as most businesses are trying to - rapidly from a monolithic structure that we have with Magento [Adobe Commerce] to a micro-service and composable layered solution, where we built a lot of microservices for our front-end web platforms, including a GraphQL front end and a React website that was interchangeable and giving us real speed and agility in our engineering.
The firm wanted to mirror this approach for its partner business, and sought a product that would wrap its order management systems services in an API technology on one efficient platform. It carried out a full RFI process to evaluate emerging vendors and the main players with a proven track record in delivering for similar businesses, along with talking to peers and looking at research, and MuleSoft came out top across all those areas.
Once the shoe designer had started engaging with MuleSoft on how it could get the order management services wrapped in APIs on their platform, the process was extremely quick. Building the first APIs happened in a matter of weeks, says Rees-John:
From signing the contract and agreeing this was the way forward we wanted to move, to getting our first dropship out with MuleSoft was a matter of about 12 weeks. The idea of utilizing APIs wasn't new to us, but the ease of how the MuleSoft platform allowed us to do it was surprising.
MuleSoft initially trained Kurt Geiger’s in-house engineers on using its platform, and the 30-strong team got to grips with the tech rapidly. The team built the APIs hand in hand with MuleSoft the first time to understand how to use the platform, according to Rees-John:
It wasn't a lift and shift and let somebody else do this for us, it was very much a self-serve operation and an upskilling of the in-house team. We don't like black box technology. We like to understand what we're trying to implement and therefore it gives us a greater understanding and frankly it's a lot more engaging for our engineering team.
Using the MuleSoft tech, Kurt Geiger created reusable APIs that connected its order management system with partner systems and reduced onboarding time for dropship partners from eight months to just 12 weeks. This was a world away from the firm’s previous tech platform, based on multiple integration technologies with different user interfaces, which was slow, prone to errors, and required a vast amount of monitoring and engineering resource. Rees-John says:
Like many businesses, we've built multiple services based out of our Magento feeds and our order management system, which was in Manhattan, and our ERP at the back end in Retail Express. We were customizing many different sources of information through that order management system and then customizing those further every time we wanted to integrate to a new partner. This was laborious and messy to try and keep a tab on changes, and when things broke, there were many points of failure we had to look at.
With the new MuleSoft tech in place, sell-through rates have improved. Having stock in one stock pot means that the firm can maximize availability for partners and distribute stock more effectively. According to Rees-John:
We can switch on and off pots of stock for different partners. We own a stock inventory pot at our end. There’s also the ability for the D2C platforms to be the first and last place things are sold. We can choose to only sell through our direct channels or we can change catalog depending on differing customer behavior based on these patterns.
Previously, Kurt Geiger didn’t have this level of control over its stock and instead had to guess consumer patterns up front at the start of the season, he adds:
You set off with the greatest intentions with a partner, but things happen. Some are much more successful than they thought they were going to be for that season, others less so. If your stock’s caught up in that situation, then you're never going to get the full sell-through from it.
The availability improvements garnered by switching from a concession to dropship model have led to a dramatic increase in generated revenue. The business is selling through more goods as availability is better for that particular consumer set.
Time saved in the engineering team has also led to a big cost saving, as the engineers aren't having to maintain old integration points, and partners are moved through the system quickly. Rees-John concludes:
We talk about 12 weeks, but it’s really the partners that tend to slow us down now more than anything else. The biggest time spent is actually in the QA and the testing, which is making sure that the finance and the stock flows are correct at the end of it.
Where a partner has got a very simple API solution at their end with a very fast platform, we have delivered all the tech changes within a matter of weeks. If we were just doing it, we could probably get a dropship up in six weeks.