How Google Demand Gen cuts e-commerce CAC by 35-45%

For most of the last decade, e-commerce brands have validated new products the same way: heavy spend on social media, fast creative iterations, and the hope that a breakout product emerges before the budget runs out.

‍

That model is breaking. In 2026, our internal benchmarks show that Cost per Mille (CPM) is reported to be up 13% year-on-year and product trend cycles are shrinking from three months to a mere three weeks. The result is that more sellers are spending more to chase shorter-lived wins, generating revenue without profit.

‍

And now, a channel that has long been under-used for this stage of the funnel has quietly become a serious alternative.

‍

Why Google, and why now

‍

The persistent industry view has been that Google Ads is a harvest channel: good for capturing existing demand, slow to ramp, and unsuitable for testing a single Stock Keeping Unit (SKU). Recent updates to Google’s Demand Gen campaign type however have changed that equation. 

‍

Demand Gen places creative-led ads across YouTube, Discover and Gmail, blending the visual storytelling that paid social is known for with Google’s intent and first-party signal. For product testing specifically, Demand Gen offers three advantages over a single-channel strategy:

‍

• Speed: Useful signals can emerge within 1-3 days of launch.
• Precision: Audiences built from search behaviour and first-party data reach consumers actively considering specific product categories. Internal benchmarks suggest customer acquisition cost (CAC) of 35-45% lower than traditional social channels.
• Longevity: YouTube and Google's broader ecosystem deliver sustained long-tail exposure, turning short-lived hits into a sustained revenue line.

‍

The three-phase framework

‍

Phase 1: Identify high-conversion products worth testing

‍

Start with what is already working organically. In GA4, plot every SKU on a two-axis matrix: traffic volume on one axis, conversion rate on the other. Products with low traffic but a strong conversion rate are the most efficient candidates for paid amplification.

‍

Structure each test at the product level, combining first-party data, lookalike audiences and custom labels for multi-dimensional testing. 

‍

Use add-to-cart (ATC) cost as the primary decision metric in the first 72 hours. A simple decision rule: 

‍

‍

Note: Always A/B test the ad creative against the landing page. Misalignment between the two is the most common reason a “winning” creative fails to convert.

‍

Phase 2: Scale the winners

‍

Around day 7, review each survivor against three metrics: return on investment (ROI), cost per acquisition (CPA), and ATC cost.

‍

‍

On top of that, continue refining the system by:

‍

• Feeding the algorithm: Upload top converting on-site search terms, customer lists, and high-intent page audiences so Google’s AI can extend reach without losing precision.
• Refreshing creative continuously: Develop new content using Google’s AI tools to fuel performance. A proven video formula is a high energy hook in the first 5 seconds, emotional resonance in the middle, and a clear CTA in the final 5 seconds.

‍

Phase 3: Optimise ROAS

‍

Once a product is scaling, return on advertising spend (ROAS) becomes the lever you tune. Break it into its components:

‍

ROAS = CVR x AOV ÷ CPC

‍

This makes optimisation diagnostic rather than guesswork.

‍

• Conversion rate (CVR): Audit the full on-site conversion journey – product page, mobile speed, checkout friction. Ad creative is rarely the bottleneck once a product has scaled.
• Average order value (AOV): Tiered offers (e.g. buy-two-get-one), bundles, and post-add-to-cart cross-sells lift revenue per order without additional acquisition cost.
• Cost per click (CPC): Bid more aggressively in lower-competition windows (e.g. late-night, weekday mornings) and tighten creative-audience alignment to lift quality scores.

‍

The takeaway

‍

This framework is about building a repeatable system in an inherently uncertain market. Using Demand Gen for product testing isn't about chasing short-term spikes. It's about directing every marketing dollar toward the products and audiences with the highest certainty of return, and giving genuine breakout products the concentrated support they need at launch.

‍