A specialty auto parts ecommerce brand with a catalog exceeding 100,000 SKUs engaged SCUBE after paid media growth plateaued. The business served a mix of retail customers, repair shops, and trade buyers, and relied heavily on paid channels to surface hard-to-find electrical components.
Performance had not collapsed. Spend was moving. But revenue had plateaued. Profit after advertising costs was inconsistent, and reporting did not reflect how demand was actually being captured across the catalog.
The objective was not short-term efficiency gains. The retailer needed a paid media system that could operate with less oversight, support catalog complexity, and scale without quietly leaking margin.
Over the 12 months following a system rebuild, profit after advertising costs increased by $530,000 year over year. Revenue grew 22% during the same period. The improvement did not come from higher bids or broader reach. It came from restructuring how demand was captured, measured, and filtered across channels.
Growth was capped by structure, not budget.
The catalog was large, uneven in performance, and treated as a single entity inside paid channels. High-intent demand existed, but signal quality made it difficult to distinguish which products were creating profit versus absorbing spend.
Measurement compounded the issue. Platform reporting emphasized surface metrics that looked healthy while obscuring margin behavior. Decisions were being made on activity, not consequence.
This combination created a familiar failure mode in large, complex catalogs. Spend scaled faster than understanding. Efficiency held just long enough to hide the problem.
The account was treated as a system design problem. The work focused on clarifying signals, separating demand types, and sequencing channels so each served a defined role.
Tracking was reconfigured to reduce distortion between platform reporting and business reality. Reporting was reoriented around profit behavior, cost of sale, revenue and channel contribution rather than vanity metrics.
The goal was not perfect attribution. It was directional clarity that could support real decisions without constant reconciliation.
Search and shopping programs were reorganized to separate branded demand from non-brand demand, exposing where growth was actually coming from.
The product feed was rebuilt to reflect performance tiers rather than completeness. High-contributing SKUs were isolated. MPNs and brand names were added to titles. Low-yield products were removed from paid paths where pricing, availability, or competitiveness made returns unlikely.
Spend began following contribution instead of volume.
Google Ads was positioned as the primary demand capture layer. Performance Max was structured to support discovery without collapsing into blended reporting. Microsoft Ads was introduced to capture incremental demand without cannibalizing existing traffic.
Across channels, ROAS functioned as a constraint. It prevented waste, but it did not define success. Stability, contribution, revenue, and profit behavior did.
Over the first 12 months:
The shift was gradual, not explosive. Performance improved as the system stopped amplifying low-value demand and began reinforcing profitable paths through the catalog.
Paid media became more predictable. Decision-making required less interpretation. The system could run without constant intervention.
“Our ROAS and Revenue are improving and we don't have wasted spend on ads that don't have a return.”
— General Manager, Specialty Auto Parts Retailer
The outcome was not a single performance spike. It was a paid media system that could scale without quietly breaking.
We offer free game plans that provide clear KPIs, campaigns insights & a 90-day roadmap.
The SCUBE Game Plan is a focused review of how large, spec-driven catalogs behave inside paid channels. It’s designed to surface what’s contributing to performance, what’s masking underlying issues, and where structure is quietly working against you.If there’s a fit, we walk through the findings in a ~60 minute conversation, looking at:
The goal is a clearer picture of how the system is behaving, so decisions stop relying on averages or assumptions.
.avif)