Case Study
Retail E-commerce Google Shopping Meta Ads

Source for Sports was spending on ads across Google and Meta. CPA was climbing every quarter and ROAS was declining. The campaigns were active - they just were not working.

Campaign audit, audience restructuring, funnel-matched creative, and product-specific landing experiences across Google Shopping, Search, and Meta - resulting in a 38% CPA reduction and 2.4x ROAS improvement within 60 days.

TL;DR
  • Source for Sports was running active paid campaigns across Google and Meta but CPA had increased quarter over quarter while ROAS declined, indicating a structural campaign problem rather than a budget or creative issue
  • A full audit revealed three compounding problems: audience targeting too broad and not segmented by intent, creative not matched to funnel stage, and high-intent traffic landing on generic category pages rather than product-specific destinations
  • Rebuilt campaign architecture from scratch across Google Shopping, Search, and Meta with audience segments matched to creative matched to landing page for each stage of the funnel
  • CPA dropped 38% and ROAS improved 2.4x within 60 days of the restructured campaigns going live
-38%
Cost per acquisition
2.4x
ROAS improvement
60
Days to results
3
Channels rebuilt: Shopping, Search, Meta

Active campaigns with declining performance almost always mean a structural problem, not a budget problem.

When CPA climbs and ROAS declines quarter over quarter while campaigns are actively running, the instinct is often to increase budget or refresh creative - but those interventions treat the symptoms rather than the cause. Source for Sports was spending consistently across Google and Meta with campaigns that were technically active and generating results, but the efficiency metrics were moving in the wrong direction over time, which is the pattern of a campaign structure that was never properly built rather than one that has been exhausted by market saturation or competitive pressure. A full audit of the campaigns, the audience structure, the creative, and the complete path from ad click to purchase was the necessary starting point before any optimization decisions were made.

The audit identified three compounding problems that were each reducing efficiency independently while reinforcing each other in ways that made the root cause difficult to isolate from within the existing structure. Audience targeting was too broad and not segmented by purchase intent, which meant budget was being spread across users at very different stages of the buying process with the same messaging and bidding strategy. Creative was not differentiated by funnel stage, so awareness-stage users were being served the same conversion-focused ads as users already in a purchase mindset, which drove up costs on clicks that were never going to convert at the rate the campaigns were bidding for. And high-intent traffic for specific product categories was landing on generic category pages that required additional navigation before reaching the relevant product, creating drop-off at exactly the point where conversion probability was highest.

Ad Examples
Creative assets coming soon

Full rebuild of campaign architecture with audience, creative, and landing experience aligned at every funnel stage.

The campaign structure was rebuilt from scratch across Google Shopping, Search, and Meta with audience segmentation as the organizing principle rather than platform or budget allocation. Users at different stages of the purchase funnel were separated into distinct audience segments with separate creative, separate bidding strategies, and separate landing destinations matched to the specific stage and intent signal of each segment. This meant each dollar of spend was working within a targeting and creative context designed for that specific audience rather than a generic campaign structure trying to serve all stages simultaneously with the same parameters.

Product-specific landing experiences were built for the top-revenue categories, replacing the generic category pages that high-intent traffic had been landing on. Each landing experience was structured around the specific product or category the user had searched for or engaged with in the ad, removing the navigation step between the ad impression and the relevant product and reducing the drop-off that had been occurring between click and purchase. Continuous optimization ran weekly against CPA and ROAS targets rather than click-through rates or impression volume, keeping the focus on the metrics that reflected actual business performance rather than platform-level engagement signals that can look healthy while revenue efficiency declines.

Ad Examples
Creative assets coming soon

38% CPA reduction and 2.4x ROAS improvement within 60 days.

With audience segmentation, funnel-matched creative, and product-specific landing experiences working together rather than independently, CPA dropped 38% and ROAS improved 2.4x within 60 days of the restructured campaigns going live. The speed of the improvement reflected the extent of the structural problems in the original setup rather than any single tactical change - when audience, creative, and landing experience are all misaligned simultaneously, fixing all three together produces compounding gains rather than incremental ones, because each improvement amplifies the effect of the others rather than operating in isolation.

Ad Examples
Creative assets coming soon

Common questions about this engagement.

How do you know when a campaign needs a structural rebuild versus incremental optimization?
When CPA and ROAS are both moving in the wrong direction over multiple quarters despite active management, the problem is almost always structural. Incremental optimization - bid adjustments, creative refreshes, budget reallocation - improves performance within an existing structure, but it cannot compensate for a fundamentally misaligned relationship between audience targeting, creative messaging, and landing experience. The signal is consistent directional decline rather than volatility or plateau.
Why does mismatched creative and landing page hurt performance so significantly?
Because the ad creates an expectation and the landing page either fulfills it or breaks it. A user who clicks a product-specific ad and lands on a generic category page has to do additional work to find what they were already searching for, and a meaningful percentage will not do that work - they will leave. The drop-off is invisible in most campaign dashboards because it looks like a landing page bounce rather than an ad click problem, which is why it often goes unaddressed for extended periods while CPA quietly climbs.
How do you segment audiences by intent without making the campaign structure unmanageable?
By starting with the two or three most meaningful intent signals for the specific business rather than trying to build comprehensive segmentation from day one. For retail, the most meaningful distinctions are usually between users who have never visited the site, users who have visited but not added to cart, and users who have added to cart but not purchased - three segments that require meaningfully different messages and bidding strategies but can be managed within a straightforward campaign structure that does not require constant complexity overhead.
What is the difference between optimizing for CPA versus ROAS in retail?
CPA optimization focuses on the cost of acquiring a customer regardless of what they spend, while ROAS optimization focuses on the revenue generated per dollar of ad spend. For a retail business with significant variation in average order value across categories, ROAS is usually the more meaningful target because it accounts for the difference between a low-value and high-value purchase in a way that CPA does not, and it aligns the algorithm's optimization objective with actual revenue generation rather than transaction volume.
Are your campaigns spending budget on the right people at the right stage of the buying process?
Most retail campaigns that are technically active but declining in efficiency have an audience segmentation problem - the budget is being allocated across users at very different stages of the purchase decision with the same bidding strategy and messaging, which inflates costs on the wrong impressions and reduces conversion rates on the right ones. Get in touch and we can identify exactly where your current structure is misallocating spend.