Talking International Expansion with Secret Sales


Secret Sales recently expanded its UK marketplace operations by launching fully localised sites in two new international markets. We sat down with Dan Ashcroft, Partner Director at Secret Sales and our MD Brett Cooper to discover why Secret Sales opted for a fully localised site strategy, what complexities this approach presented and how challenges turned into innovative solutions across many aspects of the project.

Dan, can you talk about why Secret Sales decided to internationalise?

So internationalising was a relatively straightforward decision for Secret Sales to make after the success of the business we launched two and a half years ago. Both pre and post-pandemic the business has grown at a phenomenal rate really. Taking the decision for the business to pursue a premium approach to the off-price marketplace in the fashion industry, which even two years after our initial launch, to be honest, is a market opportunity that is largely still untapped. When you think of that untapped potential, moving that model to other markets and doing so aggressively and quickly, becomes almost a necessity in order to keep that momentum growing, and to keep our Secret Sales story growing as well.

Dan, there are many ways online retailers can expand sites internationally, but Secret Sales chose to go fully localised. Why did the business decide on this strategy?

The decision to go fully localised wasn’t one that we took lightly as a business. As most ecommerce and online retailers will know it’s much easier to adopt a light-touch approach to rolling out international websites. That typically means simply providing a localised language website, often not even translating product data for users to shop on your global website. The cost of doing so is substantially lower, obviously, the effort is definitely lower and the speed is often quicker. 

So obviously, we did the opposite, we went fully localised. But when you put in the time to understand the market and the effort required to deliver a fully localised experience for potential customers, it quickly becomes one that you can understand, rationalise and build a business case around. It’s not simply about the website and the buying experience either, although that is a significant part of it, market research shows that people heavily prefer to shop in their native tongue and are more likely to trust product data, coming through marketing channels in their mother tongue, rather than in a standardised language such as English for example. 

From our perspective, where investing in localisation comes into its own, is from a culture, needs and habits perspective. Which sounds a little bit soft in terms of its factor but as a business over the last two and a half years, we’ve made a very conscious decision to maintain a concerted richness in how we propose our partners and brands to our customers. We’re not what we would call a horizontal marketplace, we are a vertical marketplace. So we do very much focus on the customer experience. 

And that’s really been one of the key pillars to our success as a business more widely. So maintaining that from an internationalisation perspective becomes pretty non-negotiable. It was really important to us, for our new customers, and that is really where localisation comes into its own. It’s about ensuring your customer is receiving the website natively, but it’s also about ensuring they feel like their unique needs and habits, within their particular country, are being met.

Brett, how did the Accelerator re-platform we architected for Secret Sales in 2019 set them up for such a large-scale project as internationalisation?

When we first set up the Accelerator for Secret Sales, it consisted of a few key systems; Akeneo for product information management, for integration across systems, Magento for the front-end ecommerce site as well as Mirakl from a marketplace fulfilment perspective. All of those systems had the ability to localise data right from the very start. And whilst it wasn’t needed in 2019, we did, particularly from a product data perspective, set up some key attributes to be localisable, even though it was localised in one language for the UK, we defined them upfront as localisable. 

That helped massively because what you tend to find is that you can’t change that after the fact in all systems. So it meant we didn’t have to completely wipe the catalogue which, for Secret Sales, is a huge catalogue and would have been somewhat problematic to do that. So having that from the start, a big bulk of the data was set up to be localisable across the systems. 

And I think the other thing is the fundamental flow of data through the system. It’s such a large catalogue now and that’s just compounded further when you add new markets as you’re just Xing the values across the different markets. So it’s quite important that you have a system that allows for effective change management. The solution is all about event-based change. So if something changes, it propagates downstream. It doesn’t send data that hasn’t changed, because frankly, the volume would be too great and you would never get the changes through in time. So that was all there from the start and I think that helped massively. And we’ve developed over the years to scale that as Secret Sales has scaled as well.

Brett, how did we approach this particular project?

Yeah, so our Accelerator re-platform solution was designed originally so it had the capability for international as I mentioned before, but what we had to do was look at the combination of brands, products and territories. So previously the solution allowed for brands and products to provide offers, a saleable SKU that someone could buy and multiple brands could place offers on the same product. You add territory to that or localised markets, and it’s just another axis that you have to use in combination. That can get quite complex when you have certain brands or partners that can sell in one market but not another, or have the ability to fulfil in different markets.

So we had to look at that whole architecture around that and expand that out. It was a proof of concept really, with the marketplace platform, with Mirakl and the systems in place, we very much worked on a proof of concept model which, I think, worked really well and fleshed out some of the unknowns that no one knew, frankly at the start. And as we progressed, we also started to look at the operational workflow with Dan and the team at Secret Sales because of the brands and partners they have, some will provide that localised content up front, but some couldn’t. So there’s a business process and a value that Secret Sales could add by having a workflow in place that enriches those products to be fully localised and finally internationalised as we talked about. That overall approach was quite important. I think it wouldn’t have been possible to predefine everything upfront, we wouldn’t have been able to do it.

Dan, do you have anything to add here?

Like what Brett just said, predefining it is almost an impossible task because actually if you look at where we were two and a half years ago, when we first re-platformed the business, we weren’t just a marketplace at the time, we were a flash sales model as well. So actually, when you take that into consideration, a mixed model for fulfilment and then two and a half years later, you are exclusively a marketplace, there are a lot more complexities in that change in direction for us as a business. So the legwork that was done beforehand, for example, putting in place localisable attributes really did save us a significant amount of effort.

But even bearing that in mind, when it came to rolling out the project, our marketplace platform, which is based on a product called Mirakl, had changed enormously in that two and a half years. To the point whereby, we were doing it as a proof of concept, but actually features were only becoming available – which helped us fulfil our requirements as an internationalised marketplace – as we were rolling out the various sprints of work. So actually what was really needed was a truly agile approach to things. And we were, very often having to re-architect things as we were going through that process. And that’s where we got, and continue to get a lot of value from Ampersand as there is that consultative approach. But actually, that ability to be agile on the fly and work out solutions as they become problems, is really where this project came into its own as we tried to meet the needs of what was quite an aggressive timeline.

Dan, Ampersand prides itself on its consultative approach and its ability to think operationally. Can you give examples of this in action?

Yeah, I think the one I’ve just given actually around needing to blend that consultation with the operational, day-to-day, is probably the best example I can give. But I’ll expand on that a little bit. That marrying of the consultative and the strategic to be fair, in our engagement to the practical hands-on attitude, really does define our relationship in so many different ways. And that’s been the case for the last three years, both on this project and on other major projects. We are talking international now, but actually, we’ve done a lot of major projects in that time, where this has been a repeated approach that is well proven for us. And that’s really been integral to the success of delivering a scalable international solution. 

To be honest, there were a number of bumps in the road in rolling this out so quickly, predominantly from our end, because we were still trying to understand exactly what we wanted to be and in what markets. But not only were we really well-advised by Ampersand, we also had a great architect and a really good project team and we were able to be flexible enough in our planning which meant when we did hit the bumps, we could elegantly work around them in quite an agile manner. 

That really forced itself towards the end of the project where, you know, timelines were becoming really important but we still had to hit the same deadlines in order to meet our conditions to roll us out to the various markets. And yet, we still managed to do that in a timeline that was right for Secret Sales in the market that we were going out to. It really was that consultative and operational approach, the blending of the two and working in an agile manner that continues to define the relationship that we have with Ampersand who we do very much see as a partner, and not as an agency or a supplier.

Brett, Ampersand has launched international sites for other retail clients before. How was this different and did it have anything to do with Secret Sales being a marketplace?

There are probably two key areas that were different with Secret Sales. One definitely linked to the marketplace, one a bit less so. So the first one was around territory-specific offers and market-specific offers for brands and partners. We hadn’t come across that before. Brand partner A can sell a product in one territory, but brand partner B sells in a different territory and then the fulfilment of that and routing of that across the entire system, was new to us, and something that we spent a lot of time working with the partners and fleshing that out at the start. 

The second one, which is linked to the marketplace, but it’s not unique to it, is the size of the catalogue. So the size of the catalogue is well into hundreds of thousands of SKUs if not in the millions I think, at the last count. So it’s quite a large catalogue, quite complex, quite varied and managing the volume of change both on the launch of a new international market where, essentially you are changing lots of products in one go, and then more from a BAU perspective, was definitely a little bit different to what we’ve seen with some of our other clients.

Dan, you recently posted on LinkedIn that this project was a “journey of addressing complex never before seen challenges”. Can you elaborate?

Genuinely it is completely true. We’ve spoken with a lot of our peers in other businesses and in some respects, they can’t quite get their heads around the things that we’ve tried to do. To provide some context, we were one of the first UK fashion businesses and one of the first UK marketplaces to expand back into the EU since post-Brexit and that in itself is a big enough challenge right now. But when you add in the complexities of doing it in a fully localised manner, and as a marketplace, it’s impossible to not look at what a daunting task is from that 50,000-foot view really. 

I’ll give you some examples based on technology but there are wider complexities around tax and logistics, in particular, which are two areas where there isn’t a well defined solution trodden at this moment in time by similar businesses or by any other industries wanting to go back into the EU and work in an online manner. Not only are these operational challenges, but they are technology ones as well. 

So, some of the examples I can give is how do you split out your order management so that you can discreetly take into consideration a number of different use cases? It might be, where is a customer based? Are they based in the UK or are they based in the Netherlands, which is one of the markets we moved into? Where is the product being sold from and shipped from? Which again has relevance in terms of the tax implication. Where is the vendor based and where are they incorporated because, again, that has an impact on the tax legislation. And finally, where is the product being shipped to? So you’ve got four different variables, which are critical in determining how you deal with that order and how you deal with that product.

Basically, you need to know scenarios. You need to understand the VAT rates for each of them and then somehow, you need to model that into your end-to-end system so that you aren’t creating a cottage industry of manual work around finance and compliance. And that was really where we had a lot of complexity. We spent a lot of time looking at financial service providers in that space, but we ended up architecting the solution ourselves with the help of Ampersand and a third party to ensure that all of our use cases and specific target markets were catered for. Not just now, but in markets that we know we’re looking at expanding into in the next two years. 

What we’ve managed to do is effectively put in place a solution that deals with all of the complexities of the orders that we are taking in the various markets, and that figures out the tax implications for each of those orders. It does it in a very succinct way that pushes the orders through our system completely autonomously without us needing to get involved in those on a case-by-case basis.

To give you some idea right now, we have a success rate of over 99% on delivery into the EU from the UK with our brands and partners. We’re really proud of that, it’s a complexity that other people haven’t really fully tackled yet. I’ve had a lot of conversations with peers in retail, who have wanted to understand our learnings on that. So, as I say, it’s a technology, operational and logistical challenge and it’s probably the most complex piece of work I’ve ever done, personally.

So Dan, this was truly pioneering then?

I mean, difficult. You know, as anybody will tell you who works in technology, it’s quite easy to take something that’s been done before, and repeat it. Especially if you have the knowledge and the expertise. And even if you don’t, you can kind of figure it out. But when you’re working on a solution that hasn’t really been figured out before, that’s the stuff that keeps you up at night.

Brett, were there any aspects of the project that worked particularly well? Any that you’d consider doing differently?

So I think what worked particularly well was the approach, without a doubt. And that’s borne out of the relationship we’ve been able to have with Secret Sales over the last three, four years. Working iteratively, working very much as a combined team that had pretty senior stakeholders and decision-makers on the ground running it, and wasn’t distilled down to more junior members of the team. We had quite a senior team on both sides looking after this. And that iterative approach, proof of concept, essentially flushing out items that we didn’t know were problems I think worked really well.

I also think, as we got ready for launch, we did a soft launch where essentially we deployed everything to production across systems but didn’t open it up to the world. And that uncovered its own issues that we hadn’t foreseen in different scenarios. So I think that worked quite well. Inevitably, there’s a bit of firefighting afterwards, as always happens with anything big, but it was quite minimal, actually, given the size and complexity of the project, because of that iterative approach and a soft launch. But I think they worked really well. 

Dan, same question?

The real benefit that we have, collectively, with Ampersand and Secret Sales is that we have a well-rounded level of experience across all of the systems in the background. Ampersand has done the front-end piece of ecommerce both in terms of internationalisation and locally many times over and that was really clear from day one, fully understanding that end to end. 

To be fair to Ampersand, where they have also picked up a lot of knowledge during the course of the last few years, has been on the marketplace piece in particular, which is complex and an area which being honest, a lot of businesses and agencies still don’t fully understand. But to be fair to Ampersand, our consultants, and our project team, they really delved into the deep end and picked up that knowledge really quickly. And we have a lot of expertise on marketplace internally, so we helped in that respect. 

And that’s really where the two sides of things came together really neatly, bringing that back-end marketplace piece into the core ecommerce solution. And actually, I’d go as far as saying that the part of this that worked really elegantly, was the piece that Ampersand was in control of, in particular the ecommerce side of it. And actually where we probably hit most of the roadblocks was more on the marketplace back-end piece. And that is because it’s not a path that’s particularly well-trodden. But collectively, we managed to get through that between not just ourselves and Ampersand, but also with Mirakl, one of our partners, and also with our other third-party stakeholders. It was something that worked really well.

In terms of lessons learned from a Secret Sales perspective, in particular, was the amount of change we continued to pursue as we were rolling out international. That was probably something that, if you speak to our wider business, our head of product and our head of technical operations, we pursued quite a large amount of business-as-usual change in combination with rolling out international. Probably in hindsight, too much so.

When you are trying to roll something out like this, you want to be in as steady of a state as you possibly can be. As a business, we do push for quite a lot of change and enhancements on our core business above and beyond just doing these projects as well. So during that period of six to nine months where we were rolling out international, we were also aggressively pursuing a strategy of continued feature development on our wider website. And while I wouldn’t say we wouldn’t do that again, actually just understanding the impact on that for a programme like this is really integral to helping you understand what you can actually achieve and what you can’t achieve.

Dan, What advice would you give other high-growth retailers or marketplaces looking at large milestone projects such as this?

From a technology perspective, for me, it would be to understand what your business can take in terms of the amount of change it wants to take on in a given period. Major projects are obviously major projects and programmes for a reason. You need to understand your own propensity as a business for change and take on the amount of change that you feel your business can take. Don’t try and do too much because actually, these kinds of changes are absolutely major in terms of what they mean, not just for your new customers and potential new markets, but for your existing customer base as well. 

If you’re going to roll out something like international, it’s obviously going to have a major impact on your business but you also have to just keep your eyes on the prize, which is your existing customer base as well as your new customer base. For me, a really critical component of this is just understanding what change your own business can take and accept and then how you build that into your own roadmap for success.

Brett, same question?

I think there are probably two things. One is the structure of the product architecture. It’s a lot harder to rethink it after you localise and move into international markets. I think looking at that upfront with your approach, whether you’re going to fully localise the product data as we did with Secret Sales, or maybe it’s a lighter touch. Being clear on that and understanding the product architecture as you’ve got it is really important. Do that first.

Two. I think you should iterate through that. What was critical for this project was having a complete end-to-end test environment. That’s not just a staging site or a test system, it’s a connected end-to-end system. It’s so critical to have that if you’re going to need to iterate because lots of the individual systems can work fine in isolation, but as soon as you connect it all together, and you put those real workflows through, that’s when things will come up. And I think being able to iterate over that is really important and definitely worth any upfront investment.

Dan, we know Secret Sales is only at the beginning of this new phase of growth but do you have any early results you can share?

So with launching into these new territories, we were able to launch, broadly speaking, around one-fifth of our existing partner and brand base into the Netherlands and Belgium, which are the target markets we’ve launched in and which is not an insignificant number of launches that you’re doing in a big bang approach. 

In addition to that we were able to, and have been since, able to launch what we call five new local heroes, which are the first sellers that Secret Sales has ever launched into those markets that are either Netherlands and/or Belgium specific, without being launched in the UK. And that’s really important because what that demonstrates is that those markets work as a standalone business, not just as part of the UK business. So we were able to demonstrate that with a number of brands that are very local to those markets and are selling and working really well. 

And then lastly and most critically, we have a large pipeline of brands and partners who want to work with us as we expand, not just into those two new international markets, but also into the new markets that are due to come across in the next two years, which is quite a robust pipeline of new launches, brands and retailers. Then I guess more on the technology side of things in terms of efficiencies and workflows, I tend to think of things as more of ‘value-add’ than looking at it from a technology and feature development, etc. front, just from my background. 

Where we’ve been able to see a huge benefit here, is working with brands and partners who haven’t been able to internationalise with other marketplaces or other retailers. The reason that has been so successful is basically 40% of our seller-base haven’t been able to internationalise with other marketplaces because they don’t have the product data in order to do that.

To Brett’s point earlier, we’ve put in place some really cutting-edge logic around how we work with product data and how we translate that data. We have a multi-faceted approach to product data that either allows us, in a workflow manner, to accept product data from our clients and partners, translated or conversely, non-translated and push that through an automated workflow, which will translate it for them. So we take away a lot of the heavy lifting for those clients and that has enabled them to move into a marketplace model. 

Similarly, from an operational and logistical perspective, we offer reverse 3PL solutions, which means that for brands that are based in the UK and are looking to ship to the EU, obviously, it’s not financially viable for individual customers to ship packages back to the UK warehouse. So again, we’ve put in solutions that enable sellers to ship back to a Secret Sales hosted facility, consign those products into pallets and then ship them back to the sellers. So in terms of what that means for our customers, again, enabling them effectively means that we have been able to tap into a large resource of sellers, brands and partners who wouldn’t otherwise be able to work in an international market.

Learn more about our Accelerator re-platform solution, how it transformed Secret Sales from a flash sales retailer into an online fashion marketplace or get in touch if you’re a high-growth retailer interested in partnering with a technical ecommerce agency that specialises in complex integration projects.