Xnurta helps Shokz reach 93% new-to-brand sales using AMC's AI-assisted SQL generation

Goals

Shokz, a consumer electronics company best known for open-ear headphones, had built a strong presence in the bone-conduction headphone category on the Amazon store in the U.S. But with that success came a new challenge: growth in their core segment was slowing, and the brand needed a strategy to keep scaling. Shokz set their sights on an adjacent category – open-ear true wireless earbuds – where they could win new-to-brand (NTB) shoppers who were not yet searching for the brand by name.

The opportunity was clear, but previous Amazon DSP campaigns that ran without advanced audience segmentation had not delivered the results Shokz needed. Brand search lift had reached roughly 9% against an 18% goal, and display click-through rate (CTR) remained below Shokz’s 0.2% goal.

To close that gap, Shokz partnered with Amazon Ads partner Xnurta to develop an insights-led audience strategy built on three integrated Amazon Ads solutions: Amazon Marketing Cloud (AMC) for audience insights, Amazon DSP for campaign activation and Amazon Brand Lift for measurement.

Approach

Xnurta’s approach centred on making AMC accessible to the broader campaign team – not just SQL specialists. To deliver on this goal, Xnurta integrated the AMC SQL Generator expert tool into their agentic solution. Available through the Amazon Ads MCP Server, which gives AI agents access to Amazon Ads capabilities, the expert tool uses Amazon’s AMC expertise to generate validated AMC SQL from natural language. For Xnurta, this made it faster to bring AMC insights to customers. For Shokz, it meant the team could translate business questions into executable AMC queries where they already work, without requiring advanced SQL skills. This reduced audience-building cycles from days to hours and gave less-technical team members the confidence to explore advanced analysis at campaign speed.

The insights unlocked through AMC journey analysis revealed where prospects dropped off between discovery and purchase. Xnurta converted these findings into four AMC audience segments: lookalikes based on high-value purchasers, category and adjacent-category keyword searchers, product detail page viewers and cart abandoners from the previous 30 days, and audiences informed by Amazon's shopper intelligence across cycling, running, business, smart-tech and wellness verticals.

These AMC audiences were activated through Amazon DSP with Performance+ optimization, which automatically adjusted delivery to reach the highest-value prospects. Generative AI supported creative variations across funnel stages, enabling Xnurta to test and refine messaging for each segment. Amazon Brand Lift measurement was implemented to quantify the upper-funnel impact of Amazon DSP exposure on brand search behaviour. The result was a faster, continuous loop from AMC insight to Amazon DSP activation to Amazon Brand Lift measurement – a repeatable model Shokz could scale for future launches and seasonal campaigns.

Results

The campaign exceeded all three primary KPIs. Amazon Brand Lift measurement showed brand search lift reached 23%, surpassing the 18% goal by 5 percentage points – demonstrating that upper-funnel Amazon DSP exposure successfully moved shoppers from passive awareness to active brand interest.¹ CTR hit 0.5%, 2.5x Shokz’s 0.2% goal, confirming that AMC-defined audiences paired with A/B-tested creatives generated stronger engagement from qualified shoppers.² And NTB sales reached 93%, exceeding the 84% goal by 9 percentage points – confirming the campaign expanded Shokz's customer base rather than simply re-engaging existing buyers.³

Beyond the headline metrics, the operational impact was significant. AMC journey-analysis and audience-build cycles that previously took days were completed in hours. Less-technical teams gained the confidence to test, refine and activate advanced audience strategies independently using AMC.

What made this campaign different wasn't any single tactic – it was the connection between them. AMC surfaced the audience insights, Amazon DSP activated them and Amazon Brand Lift measured whether awareness actually shifted behaviour. And because AI-assisted SQL generation removed the technical barrier to AMC, the full team could run that cycle – each step informing the next, getting faster with every iteration. For Shokz, connecting audience insights to activation helped them see a saturated category not as a ceiling, but as an opportunity to grow.

Sources

^1-3 Shokz, U.S. 2026.