Case Study Period: September–November 2024 | Last Updated: April 2026
Case Study | E-Commerce · Mature · International
A Leading E-Commerce Brand Didn't Win Black Friday in November. They Won It in September.
Google ShoppingPerformance MaxFeed InfrastructureNew Customer AcquisitionPeak Season StrategyInternational
Quick Answer — What This Case Study Covers
How do you scale e-commerce Non-Brand Shopping profitably at peak season?
Rebuild the feed in September — clean titles, accurate categories, supplemental feed with custom ROAS labels segmenting every SKU by performance tier
Restructure campaigns around product grade — not product category. Winners into aggressive bidding, bleeders excluded, untested SKUs into controlled discovery
Lock the feed in October — freeze all structural changes and give the algorithm 6–8 weeks to stabilize on clean data before peak
Scale bids in November — enter peak with a proven, stabilized system and scale with full algorithmic confidence
Result: +61% impressions, +72% visits, 82% new customer rate, ROAS held 40% above target — built on infrastructure work done before the season, not in-season tactics. Enter peak with a proven, stabilized system instead of adjusting bids at the last minute.
Most e-commerce brands are still adjusting bids in November. The ones that win peak season finished their infrastructure work in September.
+61%Impression growth Non-Brand Shopping YoY
+72%Visit growth Peak Season YoY
82%New customer rate Non-Brand Shopping
Client Introduction
Our client is a leading 8-to-9 figure e-commerce brand with a proven product catalog, a loyal customer base, and operations spanning North America and Europe. They had established real brand equity and genuine market demand. What they needed was the acquisition infrastructure to scale new customers profitably — at the volume and velocity their catalog size and market position demanded.
Akorn took over full Google Ads account ownership across all markets. The mandate was clear: grow Non-Brand Shopping — the channel most effective at reaching genuinely new customers — without sacrificing the profitability discipline a business at this scale requires.
IndustryE-Commerce / DTC
StageMature — 8-to-9 Figure Revenue
MarketsNorth America and Europe
SolutionsGoogle Shopping · PMax · Feed Infrastructure · International Expansion
This case study is relevant if your DTC brand is experiencing any of these
Google Shopping ROAS looks fine but new customer growth has stalled
Peak season performance is inconsistent year over year despite increased budget
Most of your Shopping conversions are coming from existing customers, not new ones
Your current agency or freelancer doesn't have a pre-peak infrastructure playbook
Your campaigns rely on Google's automation without feed-level segmentation
Challenge
At scale, Non-Brand Shopping is where most direct-to-consumer brands quietly leak budget. Without product-level control, Google's algorithm allocates spend toward whatever converts fastest — not necessarily the new customers worth acquiring at a margin the business can sustain. The result is predictable: budget flows toward the "easy wins" (returning users and branded queries), while your high-margin or strategic products remain completely underrepresented.
The challenge heading into peak season was to meaningfully grow impression share and visit volume versus the prior year while holding ROAS at or above target. With a catalog spanning multiple markets, product lines, and price tiers all competing for the same algorithmic attention, the account needed structural work — not just bid adjustments.
The deeper problem: most online retailers attempt this work too late. Feed rebuilds and campaign restructures made in October leave the algorithm insufficient time to learn before peak hits. The brands that win Black Friday typically finished their infrastructure work eight weeks earlier.
"Most online retailers are still adjusting their bids in November. The ones that win peak season built their infrastructure in September."
— Annie Lee, Google Ads Architect, Akorn Media
Strategy
The lever wasn't bids or budget. It was the feed — and the timing of when to build it. Akorn rebuilt the product data infrastructure in September, let it stabilize through October, then scaled aggressively into peak with a system the algorithm already trusted.
September
Build
Full feed audit and rebuild. Supplemental feed with custom ROAS labels deployed. Campaign architecture restructured around product performance tiers.
October
Stabilize
Feed locked. Algorithm given 6–8 weeks to learn on clean, structured data before peak. No structural changes — monitoring only.
November
Scale
Bids increased into a fully stabilized system. Algorithm entered peak with maximum signal quality and minimum volatility.
Step 01
Feed audit and rebuild. Identified attribute gaps across the full product catalog — weak product titles, miscategorized product types, missing data fields, and inconsistent pricing signals across markets. Rebuilt the Google Merchant Center (GMC) feed structure with keyword-rich titles, accurate categories, and clean attribute mapping across every SKU.
Step 02
Supplemental feed with custom ROAS labels. Deployed a supplemental feed via Google Sheets registered in Merchant Center, enabling dynamic custom label overlays by performance tier. Labels segmented every product by ROAS grade, price point, and margin proxy — giving campaigns the product-level intelligence Google's algorithm cannot generate on its own.
Step 03
Campaign architecture aligned to product grade. Restructured Non-Brand Shopping campaigns around product performance tiers — not product categories. High-performing SKUs isolated into aggressive bidding environments. Bleeders excluded from all campaigns. Untested products routed to a controlled discovery structure with spend guardrails.
Step 04
Pre-peak feed lock and stabilization. Froze all feed and structural changes six weeks before peak. Algorithm given full October to learn on clean data — no in-season disruption. The system entered peak already proven and stable.
Step 05
Confident bid scale into peak. With infrastructure stabilized and data quality proven, bids scaled into peak without algorithm disruption. The system knew exactly what to do — because we had given it six weeks to learn before the pressure hit.
Results
Peak Season Performance (Year-over-Year) — Non-Brand Shopping.
+61%
Impression growth
Non-Brand Shopping reached significantly more in-market shoppers YoY — alongside rebuilt feed data and campaign structure, without higher bids or budget.
+72%
Visit growth
Qualified, non-branded traffic at scale — the hardest type of customer for any digital storefront to reach profitably and the most valuable to acquire long-term.
40%
Above ROAS target
Held profitability comfortably above the brand's target floor throughout peak — controlled expansion without sacrificing the margin discipline the business required.
82%
New customer rate
Of all Non-Brand Shopping conversions were net new customers — consistent with what the channel was architected to do.
+175%
Shopping scale achieved in the quarter immediately following the feed rebuild. The peak season results didn't happen because of what we did in November. They happened because of what we built in September. Feed quality is the input most DTC brands ignore heading into peak — and the lever with the highest compounding return when you get it right.
Akorn Media account data, Peak Season. Full account ownership, Primary Market.
Conclusion
Across North America and Europe, the difference between a good peak season and a great one isn't what you do in November. It's what you build in September.
This brand entered peak with a fully rebuilt product data infrastructure, a campaign architecture aligned to real performance tiers, and a supplemental feed system that gave Google's algorithm the inputs it needed to find the right customers at the right cost. The result was meaningful, compounding scale — more impressions, more visits, more new customers — without sacrificing the profitability discipline the business required.
If your brand is spending on Google Shopping and your peak season results have plateaued, the problem is almost certainly upstream of your bids. It's in your feed.
The brands that win Black Friday don't react in November. They build the system that wins — in September.
Frequently Asked Questions
Questions from DTC Founders and CMOs preparing for peak season — answered directly.
Why did my Google Shopping performance plateau even though ROAS looks fine? +
A performance plateau with a healthy ROAS usually means Google's algorithm is restricted by your account structure, not your budget. The system is likely taking the path of least resistance — harvesting existing branded demand or retargeting past visitors — instead of taking risks on non-brand discovery. To break the plateau, you must segment your feed data to force the algorithm to bid aggressively on net-new customer acquisition.
How do I scale new customer acquisition through Google Shopping without destroying profitability? +
You scale profitability by segmenting your catalog by real performance tiers, not just by product category. By using a supplemental feed to isolate your proven "Hero" SKUs into aggressive bidding environments and excluding margin "Bleeders," you give Google the strict guardrails it needs to increase spend only where your margins actually allow it.
How early should I start preparing Google Shopping for Black Friday and peak season? +
You should begin your Google Merchant Center feed rebuilds in September. Google's algorithm requires 6 to 8 weeks to learn, test, and stabilize on new structural data. If you are making foundational feed changes or restructuring campaigns in late October or November, you are introducing algorithmic volatility right when you need absolute stability.
What does a Google Shopping feed rebuild actually involve and how long does it take? +
A comprehensive feed rebuild typically takes 2 to 4 weeks. It involves auditing your entire catalog, rewriting product titles to match high-intent search terms, correcting miscategorized product types, and attaching a supplemental feed with Custom Labels (such as margin grade or ROAS targets) to give the bidding algorithm deeper business intelligence.
Why is Non-Brand Shopping important for DTC brands and how is it different from branded search? +
Non-Brand Shopping captures users actively searching for a solution (e.g., "stainless steel cookware"), while Branded search captures users who already know your company name. Non-Brand is critical because it is the primary engine for net-new customer acquisition and market share growth, whereas Branded campaigns often just capture existing, bottom-of-funnel demand.
How do I know if my Google Shopping feed is hurting my campaign performance? +
If you see a high impression share on branded terms but low visibility on generic, high-intent terms, your feed is under-optimized. Other major warning signs include a small handful of SKUs eating 80% of your budget, frequent Merchant Center disapprovals, or an inability to scale your spend despite raising your daily campaign budgets.
Written by Annie Lee
Founder, Akorn Media
Over 15 years in paid acquisition. Previously worked with Gap, Amgen, Dr Squatch, AG1, and Triple Whale.
We'll audit your Google Shopping feed, grade your catalog, and show you exactly where the structural gaps are — before they cost you when it matters most.