How a €1B Luxury Fashion Brand Cut CPA by 20% in Six Weeks With Warehouse Activation

We partnered with a global luxury fashion house generating over €1B in annual revenue, with a strong heritage of craftsmanship and timeless design across ready-to-wear, accessories, footwear, fragrances, and cosmetics. With a mature data foundation already in place and Databricks established as the single source of truth, the brand was ready to take its customer activation to the next level, elevating performance while preserving the standards expected of a luxury experience.

Working alongside Jakala, the initiative focused on translating that data maturity into real-time, high-quality activation across paid media, ensuring audiences could be deployed with precision during peak campaign moments, without slowing teams down or compromising regional execution.

Overview

Key challenges:

• Manual audience uploads into ad accounts
• Wasted ad spend from poor suppression and stale data
• Fragmented campaign operations across regions and platforms
• Customer experience that didn’t meet luxury brand standards during peak activation periods

Key evaluation considerations:

• Needed to activate warehouse data across paid media
• Already invested in Databricks as the single source of truth for customer data
• A high-stakes campaign timeline required rapid execution

Solution:

• Deployed Hightouch as a composable CDP on top of Databricks
• Enabled suppression, retargeting, and lookalike audiences
• Went live in under six weeks to support a critical campaign activation window

The outcomes:

• 20% reduction in cost per action (CPA) during peak campaigns
• Six-figure savings on seven-figure ad spend
• Real-time audience activation with no manual uploads
• Improved relevance and customer experience

The challenge

Peak campaign activation periods are among the most important moments on the retail calendar, with large volumes of customer spend flowing through a short, high-stakes window. For a global luxury brand operating at scale, these moments demand precision, consistency, and speed across regions and channels.

The brand had already invested in a clear data strategy, with rich first-party customer and product data centralized in Databricks as the single source of truth. The ambition was clear: activate that intelligence across paid media to improve performance and deliver the premium experience their customers expect.

The challenge was operational. While the data foundation was strong, the brand lacked a scalable way to translate it into live audiences across platforms and regions, especially within the tight timelines of peak campaigns. Activation required multiple handoffs between teams, manual list building, and repeated uploads into each ad channel. That friction slowed execution, limited consistency, and made it difficult to keep audiences fresh during periods when performance mattered most.

This gap also impacted customer experience. Without a reliable way to suppress recent purchasers or refine targeting in real time, customers could see ads that felt irrelevant or mistimed, falling short of the brand’s luxury standards.

The same constraint limited growth initiatives like lookalike audiences. Although the brand had access to high-quality first-party data in Databricks, generating and refreshing seed audiences at the pace and scale needed for effective prospecting remained difficult. Lists had to be manually assembled, uploaded, and maintained across multiple platforms and regions, leading to incomplete or outdated inputs and reducing lookalike performance at the moments efficient acquisition mattered most.

The evaluation

Jakala played a strategic role in shaping the brand’s activation approach. As a long-term partner, Jakala had already helped the brand strengthen its data maturity by implementing a Databricks-based data lake, establishing a single source of truth and reusable models that could support customer intelligence at global scale.

With that foundation in place, the question wasn’t whether the brand had the right data strategy, it was how to operationalize it across paid media in a way that matched the brand’s pace and standards. The brand turned to Jakala to define the activation model that could extend their Databricks investment into real-time execution, especially within a high-stakes campaign timeline.

Jakala guided the brand toward a composable CDP approach, rather than a traditional CDP. A traditional CDP would have introduced long implementation cycles and forced the brand to duplicate data outside of Databricks, adding complexity and slowing down the very outcomes the team needed most. Instead, Jakala recommended a solution that could sit directly on top of Databricks, allowing the brand to activate both customer and product data across ad channels quickly and consistently, without disrupting the existing data foundation.

The solution

With the brand’s Databricks foundation already established as a trusted single source of truth, Jakala focused on one goal: turning that mature data strategy into paid media activation fast, without adding new infrastructure or slowing teams down during a high-stakes campaign window.

After evaluating composable CDP options, Jakala selected Hightouch as the activation layer on top of Databricks. This approach allowed the brand to reuse existing data models, preserve governance and privacy requirements, and move directly from warehouse intelligence to live audiences across platforms, without duplicating data into another system.

Jakala led the end-to-end orchestration, from aligning teams on the activation strategy, to modeling the right customer and product signals, to designing the audience framework for suppression, retargeting, and lookalikes. With Hightouch in place, these audiences could be continuously refreshed and deployed across regions and channels, enabling consistent execution at the exact moments where relevance and timing mattered most.

Within six weeks, the brand was operational with warehouse-powered activation in time for a critical peak campaign period. The marketing team could:

• Suppress recent purchasers to protect the premium customer experience
• Retarget high-intent shoppers based on the exact products and categories they engaged with
• Build stronger lookalike audiences using high-value customer signals drawn directly from Databricks
• Launch with confidence across multiple ad platforms, supported by a scalable operating model

The outcomes

With Hightouch in place, the marketing team could easily sync the right data into ad channels and build the audiences they needed using Hightouch’s no-code audience builder.

Smarter suppression to reduce wasted spend

By activating conversion and purchase data directly from the warehouse, the brand automatically suppressed recent purchasers and customers who had already converted on specific products. This ensured ad spend was not wasted on users who had already completed their purchase, particularly during high-volume periods when suppression delays can quickly compound inefficiencies.

More precise segmentation for targeting and seed lists

By building audiences directly from warehouse-modeled customer and product data, the team created higher-quality seed lists and more precise targeting segments, improving both acquisition efficiency and reach.

The brand’s higher-quality seed audiences, such as high-value customers with a strong affinity for handbags, provided ad platforms with clearer signals about the types of customers to prioritize. This enabled algorithms to more efficiently identify and reach similar high-value new prospects through lookalike audiences.

At the same time, the marketing team used the same warehouse data to create more precise audience segments for retargeting campaigns, based on factors such as product affinity, category interest, region, and engagement behavior.

Intent-based retargeting grounded in real behavior

High-intent shoppers were retargeted based on actual product and category interactions, enabling ads to reflect genuine customer interest rather than generic messaging. This ensured retargeting remained timely and relevant, even as customer behavior changed rapidly during the campaign period.

All these improvements, enabled by Jakala and Hightouch, resulted in a 20% reduction in CPA, equating to six-figure savings on seven-figure media spend.

The future

With a composable CDP now embedded in their data infrastructure, the brand has unlocked a range of opportunities for the marketing team. The brand is evaluating Hightouch Identity Resolution and Match Booster to improve match rates and reach in paid media, unify online and offline profiles in Databricks, and power richer CRM use cases such as win-back journeys, in-store clienteling, and omnichannel personalization.

The brand is also exploring how to use Hightouch to power conversion APIs, sending richer event data to ad platforms to help algorithms optimize campaign performance.