A blinds retailer reported 14x ROAS on streaming TV, measured independently

Blindster’s COO, Oscar Romano, wanted streaming TV to work as a sales channel, with retargeting as a secondary goal. The problem with TV has always been the same: it is easy to spend and hard to prove. A small company can’t afford to run a brand campaign on faith for six months.

It ran CTV on Vibe.co’s self-serve platform and connected the campaign to Northbeam so results were checked outside the ad platform’s own reporting. Once the numbers justified it, the team scaled. Romano’s summary: “The signals we were getting were accurate. That told us we could trust the data and act on it.”

Vibe.co reports 14x ROAS on streaming TV and a $31 cost per purchase, with budget scaled 7x within weeks of launch. After the Northbeam integration, the campaign showed a 19% ROAS efficiency improvement, an 11% reduction in customer acquisition cost and a 5% increase in average order value. Northbeam is a multi-touch attribution model, not a holdout test, so these are attributed results rather than proven incremental ones.

Measurement came first, then budget. The same sequence shows up in other published streaming TV cases. Blu Dot, a furniture designer and retailer, connected Roku Ads Manager to its Shopify store so TV views could be linked to purchases; around its annual sale, purchase conversions went from nine a day to 50, the company raised its investment 10x, and Roku reports a 2,308% return on ad spend, about 23x, over an 11-week campaign period. Doheny’s, a pool supply retailer, reported 5.46x higher ROAS year over year after switching from standard CTV audience segments to MNTN’s matched, intent-based audiences. None of the three case studies reports a randomized incrementality test; every figure is reported by the platform or the attribution tool. That is exactly why an independent measurement layer matters.