Moda Costa Digital increases online sales by 180% with abemonFLOW

The challenge

Moda Costa Digital is a fashion company based in Marbella that operates 6 physical stores along the Costa del Sol, a direct-to-consumer online store, and presence on 3 marketplaces (Amazon, Zalando, El Corte Ingles). Founded as a Mediterranean fashion brand focused on linen, organic cotton, and sustainable textiles, the company had grown to 12 million euros in annual revenue. But that growth had created a problem threatening to undermine the brand: three sales channels operating as independent companies.

The point-of-sale system for physical stores was one system. The eCommerce platform (Shopify) was another. And each marketplace had its own management panel. Inventory wasn’t synchronized in real time: an item sold on Amazon could continue showing as available on the website for hours, generating overselling and cancellations that frustrated customers. A garment returned in-store that had been purchased online required a 15-minute manual process because the systems didn’t share information.

The return rate was the most acute pain point. eCommerce had a 28% return rate, significantly above the fashion online average (22%). Internal analysis revealed that the primary reason wasn’t product quality but sizing issues: customers didn’t know which size to order, so they ordered two or three, kept one, and returned the rest. Each return cost the company an average of 8.50 euros in reverse logistics, inspection, repackaging, and stock reintegration. With 14,000 annual returns, the cost exceeded 120,000 euros.

Customer service was another friction point. A team of 5 people handled inquiries via email, website chat, Instagram DMs, and marketplace messages. 70% of inquiries were repetitive: order status, return policy, sizing guide, stock availability. But each inquiry required the agent to search for information in the system corresponding to the originating channel, which multiplied response times. The average first response time was 6 hours — unacceptable for customers expecting immediacy.

The solution

The 4-week Blueprint identified three root problems feeding all the symptoms: disconnected channels with no single source of truth, absence of a unified customer profile, and customer service processes that depended on manual searches across multiple systems.

We implemented four Engine layers, with a particular focus on end-customer experience:

Integration unified the channels. We connected Shopify (eCommerce), the physical store POS, Amazon Seller Central, Zalando Partner Portal, and El Corte Ingles Marketplace into a centralized data hub. Inventory synchronizes in real time with under 30 seconds of latency: when a unit sells on any channel, availability updates instantly across all others. Returns follow the same principle: an in-store return of an online purchase automatically updates the order status and reintegrates the garment into available stock.

Data built the unified customer profile that didn’t exist. A customer who shops in-store, on the website, and on Amazon is now a single entity with a complete history: sizes purchased, sizes returned, preferred categories, purchase frequency, and customer lifetime value. This profile feeds the size recommendation system that became the project’s most impactful change. The system analyzes the sizes the customer has purchased and kept (not the ones they returned), cross-references with the specific measurements of each garment in the catalog, and generates a personalized recommendation with a confidence level. On product pages, the customer sees a message like “Based on your previous purchases, we recommend size M with 92% confidence.”

Orchestration automated order, return, and customer service workflows. The order flow is now omnichannel end to end: a customer can buy online and pick up in-store, buy in-store and receive at home (for sizes or colors not available), or return at any location regardless of purchase channel. Each flow has its automated workflow with corresponding notifications, stock updates, and accounting entries. The customer service chatbot, powered by OpenAI, resolves 60% of inquiries without human intervention: order status, personalized sizing guide, return initiation, store availability, and product recommendations. Inquiries it can’t resolve are escalated to the team with all context already gathered.

Design was the layer that directly touched the customer experience on the web. We redesigned product pages with the size recommendation visible, photos with models of different sizes for visual reference, and a simplified purchase flow that reduces friction. The website search was enhanced with semantic search: the customer can type “blue dress for summer wedding” and get relevant results, not just exact keyword matches. The mobile experience, which represented 72% of traffic, was optimized with a 3-step checkout.

The results

Online sales grew by 180% in the first 8 months. The growth is attributed to three factors: size recommendations reduced purchase friction (customers buy with more confidence when they know the size will be right), semantic search improved homepage conversion by 45%, and the omnichannel experience generated cross-channel sales that were previously impossible. 22% of customers who visit a physical store and don’t find their size complete the purchase online the same day thanks to the unified profile.

Returns were reduced by 40%, dropping from 28% to 17%. Personalized size recommendations were the primary factor: customers no longer order multiple sizes “just in case.” The direct savings in reverse logistics exceeds 50,000 euros annually, but the real impact is greater because every avoided return is a net sale that was previously a cost.

Project ROI reached 3.5x in 8 months. The total project investment was recovered through the sales increase and return reduction. The CEO sums it up with a metric she likes to use in investor meetings: the project cost was equivalent to 2 months of returns under the old model.

Customer service automation reached 60%. The chatbot handles order status inquiries, size recommendations, return initiation, and stock availability without human intervention. The team of 5 was reduced to 3, not through layoffs but reassignment: two people moved to the digital marketing team where their customer knowledge is more valuable. Average first response time dropped from 6 hours to 45 seconds for automated inquiries.

Engine layers used

• Integration: Omnichannel hub connecting Shopify, POS, Amazon, Zalando, and El Corte Ingles with real-time inventory sync and unified return management
• Data: Unified cross-channel customer profile, size and preference history, personalized size recommendation model with confidence level
• Orchestration: Omnichannel order and return workflows, OpenAI-powered customer service chatbot, contextual escalation to human team
• Design: Product pages with size recommendations, semantic search, optimized 3-step mobile checkout