How we scaled CPAP.com's paid media with Profit-on-Ad-Spend (PoAS) bidding

Placeholder copy — this case study has no source document yet. Draft to be provided by the CPAP.com stakeholder before publish.

Context

CPAP.com is a large US e-commerce retailer in the sleep-therapy category. They asked us to help move their paid media programme beyond ROAS-based bidding — because ROAS alone was systematically over-weighting high-revenue SKUs with thin margins and under-weighting genuinely profitable products.

Challenges

• ROAS-only Smart Bidding in Google Ads was scaling low-margin campaigns.
• No margin data was reaching the ad platforms — bidding was blind to profitability.
• Product-level margin lived in an internal system, not in the ad-platform conversion payloads.

Solution

• Built a nightly margin feed in BigQuery joining orders, cost-of-goods, and refunds.
• Piped item-level profit into GA4 as a custom parameter on purchase.
• Routed the profit-weighted conversion into Google Ads via server-side GTM, using it as the primary bidding signal.
• Mirrored the same signal into Meta CAPI for parity across platforms.

Results

To be filled with real numbers once the reporting window closes.