How push ads scaled eCommerce campaigns to $833k with consistent ROI

eCommerce marketers often need to balance rapid growth with sustainable returns. Customer acquisition costs continue to climb across major platforms, while consumer expectations for personalization and speed make campaigns harder to scale profitably. This means that finding cost-efficient and repeatable traffic sources is essential.

This article examines how one media buying team generated more than $833K in eCommerce sales while maintaining a stable 20–30% return on investment. The campaigns were built around push notification ads, combined with a disciplined approach to testing, scaling, and ongoing optimization. Optimization strategies were offered by Sotos Charalambous, PropellerAds Account Manager.

Rather than a one-time success, the results highlight how structured processes and careful traffic management can deliver consistent profitability across multiple projects. The following sections outline the campaign setup, performance data, optimization strategies, and lessons that other marketers can apply to their own eCommerce growth efforts.

Background and campaign setup

The advertiser behind these campaigns managed two distinct eCommerce projects. The first (Project A) focused on a global online marketplace offering a wide range of products, while the second (Project B) centered on a fast-fashion retailer targeting younger audiences in Western markets. Both projects relied on a commission-based affiliate model, where revenue was tied directly to completed sales.

To manage acquisition costs, the team used CPA Goal bidding, which automatically adjusted bids toward a target cost per action. Conversions were tracked through postback integrations, allowing real-time feedback between spend and performance.

After a range of tests, Push notifications were selected as the primary ad format. For eCommerce campaigns, Push remains a valuable channel: it provides immediate visibility on mobile devices, scales across multiple regions, and supports rapid testing of creatives and landing pages. The campaigns also incorporated other formats such as In-Page Push and Survey Exit, but Push consistently delivered the highest volume and strongest return.

Testing and scaling framework

The campaigns followed a phased lifecycle designed to minimize early risk and build momentum gradually.

Warm-up period: Campaigns began at least one week before major shopping events such as Singles’ Day, Ramadan, or Black Friday. This allowed the algorithms to gather data ahead of peak demand and reduced the volatility of last-minute launches.

Days 1–7 (Testing): During the first week, the focus was on broad data collection. The team ran A/B tests across landing pages, GEOs, and creatives, while monitoring zone-level performance. Early conversions often appeared with a delay, reinforcing the need to avoid premature decisions and give the data time to stabilize.

- Advertisement -

Days 8–20 (Scaling): Once baseline performance was established, budgets were gradually increased for high-performing zones. Underperforming sources were removed, and CPA targets were adjusted dynamically to sustain profitability.

Ongoing optimization: After the initial scaling, campaigns moved into continuous refinement. Every two weeks, zones with low conversion rates were excluded. Monthly reports also flagged traffic sources with high cancellation rates, ensuring budgets were consistently directed toward the most reliable segments.

This step-by-step approach created a balance between scale and efficiency, turning testing insights into sustainable growth.

Results and performance data

Project A: Global marketplace

This Push campaign ran for more than a year and generated over $833K in total sales. Average ROI remained above 20%, with significant month-to-month variations. April 2025 was a standout, with nearly $99K in revenue on $30K spend — a 225% ROI. Other months, such as January 2025 (17% ROI) and December 2024 (25% ROI), showed more moderate but steady performance.

Top GEOs for this campaign included: MX, FR, ES, IL, IT, CL, DE, BR, SA, KR, CO, PE, GH, NL.

The marketplace’s broad product range attracted a high-volume, price-sensitive user base, especially in emerging markets. Android devices dominated in these regions, reinforcing the importance of platform-level targeting in eCommerce campaigns.

Project B: Fast fashion retailer

The second project ran for 4-5 months and produced strong results across iOS-focused Tier-1 markets. Besides Push, In-Page Push for iOS, and Survey Exit format were used. Average ROI reached 39.7%, supported by high engagement among Gen Z and millennial shoppers who responded well to trend-driven promotions and limited-time offers.

In this case, iOS traffic consistently outperformed Android, reflecting higher conversion rates in premium markets. The contrast with Project A highlighted how device distribution and purchasing power can significantly influence campaign outcomes.

Together, the two projects demonstrate that sustained investment in Push, In-Page Push and Survey Exit traffic, supported by careful optimization, can produce both scale and profitability across different eCommerce offers.

Optimization strategies that worked

The results were achieved through systematic optimization rather than one-off tactics. Several practices proved particularly effective:

Zone management

The team conducted bi-weekly performance audits to identify underperforming traffic zones. Zones with persistently low conversion rates or high order cancellation rates were blacklisted. This continuous pruning process ensured budgets were redirected to reliable sources.

Dynamic bidding

Instead of fixed bids, the campaigns relied on dynamic CPA bidding. By passing payout data through the postback, bids were automatically adjusted in real time according to the value of each conversion. This approach allowed higher bids for profitable sales while keeping spend proportional to value.

Seasonal timing

Campaigns were aligned with major shopping events such as Singles’ Day, Ramadan, and Black Friday. Launching one week in advance gave algorithms time to adapt before traffic spikes, improving efficiency during peak demand.

Landing page testing

Multiple landing page variations were rotated during the testing phase, ranging from default product pages to seasonal sales links. This allowed the strongest performers to emerge based on conversion data, rather than assumptions.

Landing page testing

Source: PropellerAds

Audience segmentation

While language targeting was not required, device and GEO segmentation proved critical. Android traffic performed best in price-sensitive markets, while iOS conversions were stronger in higher-income regions. Creative selection, bidding, and landing pages were adjusted accordingly.

These strategies created a feedback loop where campaign health was preserved, profitability maintained, and scaling opportunities maximized.

Broader ecommerce lessons

Several takeaways from these campaigns extend beyond the specifics of push traffic.

Data patience matters

Conversions did not always appear immediately. In the first few days, results often lagged, underscoring the importance of allowing data to stabilize before making changes. Acting too quickly could have led to cutting profitable zones prematurely.

Device targeting reflects consumer behavior

Android performed better in emerging, budget-conscious markets, while iOS delivered higher conversion rates in premium regions. Aligning campaign setups with these patterns was essential to sustaining ROI.

Influencer activity complements performance ads

In fashion ecommerce especially, influencer visibility created demand that performance campaigns could then capture. Recognizing how earned attention interacts with paid channels helped maximize results.

Gamification is an emerging growth lever

Campaign managers observed that some ecommerce platforms were experimenting with gamified shopping experiences, such as rewards and bonus systems. For performance marketers, understanding how these dynamics affect user engagement may open new opportunities in future campaigns.

Final takeaways

These campaigns demonstrate that ecommerce advertisers can achieve both scale and sustainability with the right mix of traffic, bidding models, and optimization processes. By generating more than $833K in sales at an average 20–30% ROI, the advertiser showed that push notifications remain a reliable acquisition channel when paired with structured testing and disciplined scaling.

The broader lesson is that long-term profitability does not come from short-lived wins but from continuous refinement; testing landing pages, aligning with seasonal demand, managing zones, and tailoring strategies to device and market conditions.

Do you want to repeat this success or do even better? Join PropellerAds for high quality traffic, efficient tools, and expert guidance.