Data-Driven ROAS Optimization

Background & Problem

Upon joining OOA Inc. in mid-August, the brand's ROAS was stagnating around 120%. The core problem was an intuition-based budgeting structure spread across multiple channels — Search Ads, Display Ads, and Instagram — without any data-driven basis for allocation. Beyond the structural inefficiency, the product messaging was written in supplier-facing language (e.g. "fusing bra") that failed to resonate with actual customers.

Strategy

The strategy had two tracks running in parallel. First, I crawled customer review data to extract consumer-language pain points — identifying "no-wire" and "side coverage" as the real purchase drivers, replacing supplier jargon with terms customers actually used. Second, I analyzed channel-level inflow and conversion data to make the case for full consolidation into Instagram. Leadership initially pushed back on the single-channel risk, so I built a linear regression model on historical spend-to-revenue data per channel, generating scenario projections that quantified the upside of consolidation. The numbers made the argument; the strategy was approved.

Execution

• Customer Review Analysis: Crawled product reviews to extract consumer-language keywords — surfacing "no-wire" and "side coverage" as the dominant purchase motivators, replacing supplier-centric product descriptions in all campaign copy.
• Channel Inflow Analysis: Analyzed conversion rates and cost-per-acquisition by channel to identify Instagram as the dominant high-efficiency source and build the data case for consolidation.
• Scenario Modeling: Built a linear regression model on historical spend-to-revenue data per channel, generating projected ROAS scenarios under full Instagram consolidation to quantify the upside and address leadership concerns.
• Tracking Infrastructure: Implemented UTM parameters and Cigro integration to ensure accurate attribution data across all active channels throughout the transition.

My Role

As a Performance Marketing Intern, I took full ownership of the analysis and channel strategy. I crawled and analyzed customer review data to reframe product messaging, diagnosed channel-level conversion inefficiencies, built the scenario model used to secure leadership approval, and established the tracking infrastructure to monitor performance throughout execution.

Results

ROAS growth — Aug to Dec 2023

Budget allocation — before vs. after

• ROAS Growth: Increased ROAS from a stagnating 120% baseline to 640% — a 5.3× overall increase — through channel consolidation and insight-driven messaging.
• Leadership Alignment: Scenario model converted initial executive resistance into full strategic approval, demonstrating that data-backed arguments can move organizational decisions.
• Business Impact: Secured structural profitability through optimized ad spend and consumer-language campaign messaging.