Simon Hetland helps apparel retailer Naked Copenhagen optimise and scale, driving digital sales performance and channels to ensure profitable growth.
Imagine this: You're an e-commerce brand, growing fast, with revenue up 77% year-on-year.
Sounds great, right? But what if, underneath that top line, your ad spend has increased by 2.5X in the same period.
That's exactly the challenge we faced at Naked Copenhagen, a global destination for women's sneakers and streetwear, just a few months ago.
The challenge: Ad spend on empty shelves
As a fashion company, we operate on a unique rhythm. We place bets on seasonal products six to nine months in advance, without knowing the exact demand for them when they launch. When a product is sold out, it’s gone. We can't just call Nike and order 2,000 more of a popular sneaker.
We have been using Google’s Performance Max — an ads campaign type that uses AI to help maximise performance across all channels — which was fantastic for scaling without expanding our internal team.
Although we were performing well, we realised we had set up our automation to prioritise products to get the highest return on ad spend (ROAS). These were often our fastest-selling items, the ones already flying off the shelves. However, our slow-moving inventory, the products we needed to shift, lagged behind.
We knew we had to shift our strategy, especially with peak retail season on the horizon.
The journey: Building an AI-enabled growth engine
We took this opportunity to our agency, HVG Media. They didn't just restructure our ad accounts. They proposed building an AI-driven growth engine that would integrate our real-time warehouse data directly into Performance Max.
It was an ambitious, bespoke project, and quite frankly, a big risk considering Google Ads touches up to 50% of our turnover.
Our collaboration was very close. We had a phased rollout over four months to ensure we protected the business from any technical glitches or hiccups.
We also ensured all comms-lines were open and that HVG Media had direct access to our CFO for economic insights and our IT team for data integration, for instance. This was crucial for building confidence at every step.
The solution: Real-time data meets Google ads
“We started by unifying data from Naked Copenhagen’s Google Ads, Merchant Center, and Google Analytics 4 (GA4),” explains Matej Benes, cloud and AI engineer at HVG Media. “Having this data together in Google Cloud helped us make product-level predictions about seasonality, price competitiveness, and better assess the business impact of each click.”
Data like real-time stock levels, size availability, sales velocity, and forecasted sell-out dates were then processed by a custom Google Cloud algorithm. This aligned ad spend with real-time inventory levels, our business priorities, and other signals the algorithm gathered from connected platforms, allowing for smarter bidding.
A crucial, and custom, metric we use is ‘weeks in stock’. This refers to how many weeks in stock the item has left before it's sold out, given the stock level and sales velocity for that product.
Damodar Aros Mardones, CEO and founder at HVG Media, adds: "So if the item has 10 weeks in stock, we can still allocate budget to the item because the stock is sufficient to last for 10 weeks at the current pace. However if this metric is e.g. 2, then we would be completely sold out within 2 weeks given the current pace, so we start to reduce or pause marketing to ensure we don't lose margin on lowering conversion rate due to availability.”
Depending on the remaining weeks in stock, the system either increases or decreases ad spend on that product, while also considering ROAS and other metrics.
For instance, if a popular New Balance sneaker in a specific size was nearly sold out, the system would automatically reduce bids for that exact SKU. Then redirect budget to other available sizes or alternative high-margin, slower-moving styles.
Conversely, if a new collection of Adidas trainers had just arrived and showed high initial demand, our system could intelligently increase bids to meet that immediate opportunity before stock depleted. It could then pivot focus once a threshold was met.
This level of real-time, intelligent reallocation was simply not possible before.
The results: Unlocking revenue growth and efficiency
Our ad spend grew by 57% year-over-year and our revenue increased 92% year-over-year following the launch of this new way of working. A stark contrast to our previous 2.5X ad spend growth for 77% revenue growth we began with. Conversion rate growth also saw a significant leap, reaching over 2X year-over-year (compared to 2.4% previously).
While the performance increases were welcome, what we were actually looking for was that we could see that the products we had the most stock on, were the ones we were spending the most on. And it did just that. Allowing better stock rotation and management for our warehouse.
We also found the impact extended beyond the measurable KPIs. This new approach actually solved another business challenge: an unsustainable mix of paid and organic traffic. By prioritising ad spend on products that needed a push, we could use other channels, like newsletters and social media, for popular items. This allowed the already popular items to sell through our organic channels, unlocking more advertising budget for the less popular items.
Finally, our internal team has also grown, creating a job opportunity within Naked Copenhagen to manage the new system.
What’s next: Black Friday and peak season
This AI-driven system has changed our approach to critical sales periods, like Black Friday. Last year, because of the way we had set up our ads automation, we often held back on spending more during Black Friday, because we couldn't control if we were spending on the right products or if it was truly profitable. We only discovered wasted potential by analysing the numbers afterwards.
Now, we have the confidence and the tool to truly scale our budget during these short, intense periods. We can make real-time decisions, knowing we’re always prioritising the products that make business sense, adapting to unpredictable demand spikes and rapidly changing inventory. This system is built for any season, giving us the agility to unlock our full potential, even when the market is at its most dynamic.
Our journey isn't over. Our next step is to integrate profitability data into our automation. We want to move beyond just stock volume and ensure that every order generated is within a profitable scope, factoring in elements like product return rates.
This project with HVG Media has redefined what's possible in scaling digital advertising profitably for us. It’s a testament to how data, technology, and tech solutions can revolutionise paid media performance for modern retailers. For us at Naked Copenhagen, it's not just about more sales; it's about smarter, more sustainable, and more profitable growth.
