Ecommerce performance marketing is not about running ads. It is about engineering a paid acquisition system where every rupee spent returns a predictable, profitable multiple. Most brands have campaigns. Few have systems. The difference is campaign architecture that separates funnel stages, creative strategy built around the ICP's actual purchase motivations, and attribution accuracy that reflects reality rather than platform optimism. Oddtusk builds that system across Meta Ads and Google Ads for ecommerce and D2C brands that need paid media to be a reliable growth lever.
- Meta Ads (Facebook and Instagram)
- Google Search and Shopping Ads
- Performance Max Campaigns
- Dynamic Product Ads
- Creative Strategy and Testing
- Feed Optimisation (GMC)
- Attribution and Tracking Architecture
Platform-reported ROAS is not revenue. The gap between Meta's claims and your bank account is where most ecommerce growth budgets are lost.
Meta and Google report ROAS using attribution models that overcredit their platforms, creating optimism that fails when compared with actual revenue. Oddtusk reconciles these figures with GA4-attributed revenue each month. We make budget decisions based on reconciled numbers, not the platform's preferred version of truth. For D2C brands, this accuracy is the foundation of every scaling decision across the full ecommerce marketing stack.
Meta Ads for Ecommerce
We build Meta accounts around funnel stages: prospecting for cold audiences, retargeting engaged visitors with social proof, and Dynamic Product Ads for cart viewers. Audience architecture is built from the ICP outward. Budgets are set according to the contribution each stage generates, so scaling decisions are grounded in lasting unit economics tracked in GA4.
Google Shopping and Search
Google Shopping is primarily a feed-quality problem, not a bidding problem. We optimise product titles and attributes in the Merchant Center before restructuring campaigns. A well-optimised feed generates significantly more volume than bidding changes on a poor structure. Search campaigns target high-intent queries with exact match keyword architectures and systematically maintained negative keyword lists.
Performance Creative Strategy
Creative is the highest-leverage variable in Meta's performance. The same budget produces vastly different results depending on whether the creative addresses purchase motivations. We produce briefs for every campaign stage based on ICP research, systematically testing UGC frameworks, product demonstrations, and urgency mechanics. Top-performing creator content is cross-amplified with consumer behaviour data.
Dynamic Product Ads Architecture
Dynamic Product Ads serve the specific product a visitor viewed, personalising creative from the catalogue feed. We configure DPA campaigns on Meta and Google with audience segmentation: viewers and abandoners receive messaging specific to their stage. Catalogue quality including pricing and image dimensions is fixed before activation to ensure every impression serves a valid, purchasable product.
Attribution and Tracking Architecture
Every engagement begins with attribution architecture. The Meta Conversions API is implemented server-side via GTM to improve signal quality post-iOS 14. Google Enhanced Conversions is configured to pass first-party data. UTM parameters are standardised across all channels. Monthly attribution reconciliation compares platform-reported and GA4-attributed revenue to identify misattribution before it drives costly scaling decisions.
Account Audit and Restructure
Most inherited accounts have structural problems: campaign proliferation, overlapping audiences, and exhausted creative that limit performance regardless of optimisation. We conduct systematic audits identifying these issues before recommending a restructure. Changes are sequenced to avoid disrupting learning periods of campaigns already performing, ensuring a smooth transition into a more efficient paid media system connected to the full ecommerce stack.
CAC down 35% after restructure. Shopping revenue doubled from feed optimisation. Actual ROAS 40% lower than platform dashboards claimed. All fixed.
Ecommerce accounts restructured with proper funnel-stage separation, audience exclusion logic, and negative keyword governance consistently reduce blended CAC by 30 to 40 percent within 60 days. All tracked in GA4 against real revenue, not platform dashboards.
Google Shopping campaigns on well-optimised feeds with keyword-priority title structures and resolved Merchant Center disapprovals consistently generate two times the revenue of the same budget on a poorly structured feed.
Brands that implement proper attribution architecture via GTM consistently find their actual contribution-positive ROAS is 30 to 50 percent lower than platform dashboards report, enabling accurate budget allocation rather than scaling campaigns that appear profitable but are not.
From attribution audit to profitable paid acquisition at scale
Audit and Baseline
Tracking Build
Campaign Architecture Restructure
Feed Optimisation and Creative Brief
Launch and Learning Period
Reporting and Scaling
Straight answers to the questions that matter.
Performance marketing for ecommerce is paid advertising managed and optimised against revenue outcomes including ROAS, CAC, and conversion rate, rather than awareness or engagement metrics. Every campaign decision is grounded in what it costs to acquire a customer and what that customer generates in revenue. For ecommerce and D2C brands, it spans Meta Ads, Google Ads and Shopping, and YouTube, all structured around the product catalogue, the customer acquisition funnel, and the unit economics that determine whether spending more produces profitable growth.
ROAS targets vary by product margin, category, and business model. Rather than chasing a benchmark ROAS, we calculate the break-even ROAS for each brand based on their actual cost of goods, shipping, and operating costs, then set ROAS targets per campaign type. Platform-reported ROAS is reconciled against GA4 attribution monthly to prevent over-reporting on Meta and Google.
We structure Meta Ads ecommerce campaigns across three distinct layers. The prospecting layer uses broad or interest-based targeting to reach new ICP audiences. The consideration retargeting layer targets website visitors who engaged but did not purchase, with social proof and product-specific messaging. The Dynamic Product Ads layer serves personalised product ads to cart abandoners based on catalogue feed data and pixel event tracking, all measured in GA4.
Performance Max is Google's automated campaign type distributing ads across Search, Shopping, Display, YouTube, Gmail, and Discover. For ecommerce brands, Performance Max works best when the pixel has sufficient conversion data: typically a minimum of 30 to 50 purchases per month. Without adequate conversion data, Performance Max defaults to impression-heavy placements that generate traffic without purchases. We run Performance Max alongside dedicated Shopping campaigns rather than replacing Shopping entirely, using asset group segmentation.
iOS 14 App Tracking Transparency reduced Meta's ability to track web conversions from iPhone users who opted out of tracking. We address this through three implementations: Meta Conversions API configured server-side via GTM to pass events directly from the website server to Meta; verified domain configuration in Meta Business Manager; and eight-event pixel priority setup. First-party data strategies including email list custom audiences are prioritised as a reliable signal source independent of browser tracking.
A Google Shopping feed is the product data file submitted to Google Merchant Center that drives what appears in Shopping ads. The quality of the feed directly determines Shopping campaign performance because Google's algorithm uses feed data to match products to search queries. We optimise Shopping feeds from the attribute level: rewriting titles with keyword-priority structure, adding missing attributes, resolving Merchant Center disapprovals, and improving image quality, before bidding strategy changes are made.
Accurate attribution requires a multi-layer approach. GA4 is configured with complete ecommerce event tracking and UTM parameters standardised across every paid channel via GTM. Platform-reported ROAS from Meta and Google is reconciled against GA4 attributed revenue monthly, with discrepancies above 20 percent triggering an attribution audit. Meta Conversions API and Google Enhanced Conversions are implemented to improve signal quality.
Creative fatigue occurs when an ad has been shown to the same audience so many times that engagement rate drops, signalling the algorithm to reduce delivery or increase CPM. On Meta, we monitor frequency per ad set and CTR trends weekly. When frequency exceeds five to seven impressions per week or CTR declines by more than 30 percent from the campaign baseline, new creative variants are introduced before performance degrades further. We maintain a creative testing calendar producing at least two to three new ad concepts per month.
For Meta Ads, a minimum of 50,000 to 1,00,000 rupees per month generates enough data for meaningful campaign optimisation across prospecting and retargeting. For Google Shopping, a minimum of 30,000 to 50,000 rupees per month provides sufficient impression volume for bid strategy learning. Below these thresholds, campaign learning periods extend significantly and optimisation decisions are based on insufficient data. We recommend starting with a single platform where the ICP is most concentrated.
Monthly reports cover platform-reported and GA4-reconciled ROAS by campaign type and audience segment, CAC by channel and campaign, conversion rate by traffic source and device, ad spend by platform with week-over-week trend, top-performing creative by CTR and conversion rate, and Shopping campaign performance at the product and category level. All figures are presented as both platform-reported and GA4-reconciled.