Free Growth Diagnostic·$1,500 Value
All Case Studies
8-Figure Mobile Accessories Brand
Google Ads · Account Rebuild

Mobile Accessories Brand. Feed rebuild and campaign consolidation. 2.59× to 4.14× ROAS.

2.59× 4.14×
Blended ROAS · Before: ~$160K/mo avg · After: ~$229K/mo avg
+60% ROAS Improvement
20+ → 3 Campaign Tiers
Week 1 Feed Impact Visible
Jan 1 – Oct 31, 2025 · Previous Management
2.59× ROAS
$1.6M Managed Spend
Google Ads account under previous management showing 2.59× blended ROAS, Jan–Oct 2025. Campaign names redacted per NDA.

Jan 1 – Oct 31, 2025 · Previous management · 2.59× blended ROASClient identity and campaign names withheld per NDA.

Nov 1, 2025 – May 31, 2026 · MESAscale
4.14× ROAS
$6.6M Tracked Revenue $1.6M Managed Spend
Google Ads account under MESAscale management showing 4.14× blended ROAS, Nov 2025–May 2026. Campaign names redacted per NDA.

Nov 1, 2025 – May 31, 2026 · MESAscale · 4.14× blended ROASClient identity and campaign names withheld per NDA.

The feed titles were SKU codes and internal model references. The query data showed low impression share on high-intent product searches: the searches that include model name, compatibility, and finish. Those were the searches we needed to be on.

More than 20 campaigns had accumulated over years, with no defined boundaries between them. Brand queries were triggering general campaigns. Shopping and Performance Max were eligible for the same product searches. At $1.6M in spend, that complexity was no longer easy to absorb.

The sequence mattered: feed first, then campaign structure, then negative-keyword governance, then audience signals, then bid targets. Over seven months, the account moved from 2.59× on the prior ten-month period to 4.14× on the same spend. $6.6M in tracked revenue.

An overgrown campaign list on top of a feed built for the warehouse

What we found when we got into the account

01

Product titles written for the warehouse, not the buyer

SKU codes, model references, internal naming: none of it matched what buyers type into Google. High-intent searches weren't triggering the right products. Lower-intent queries were.

02

Multiple campaign types eligible for the same demand

With 20+ campaigns and no query boundaries between them, Shopping, Search, and Performance Max were all eligible for the same searches. Budget distributed itself without any signal about which campaign was the right fit.

03

Performance Max launched without audience signals

PMax had no Customer Match data connected. No purchase history, no CRM data. The algorithm started from scratch on every query. The client had a full, segmentable customer list that had never been linked to the ad account.

04

ROAS targets set to aspirational numbers, not account history

Bid targets were set to what the client wanted to hit, not what the account had been doing. Chasing an unreachable number restricted spend, reduced impressions, reduced conversions, and starved the algorithm of data.

Free Growth Diagnostic  ·  No Pitch  ·  Yours to Keep

If you recognized your account in those findings, a diagnostic session surfaces the same structural gaps and gives you a prioritized plan to address them.

Apply for Free →

From 20+ overlapping campaigns to three defined roles.

Three layers, defined query boundaries. Brand Defense owns branded search. Core Shopping handles structured product demand. Performance Max handles broader discovery, anchored by audience signals. Budget flows to the right layer for each query type.

Before  ·  Account Audit: Campaign Inventory
# Campaign Type Issue Identified
01 Brand KW Search Overlap with Search Gen on branded queries
02 Search Brand Search Duplicate branded coverage, no negative wall
03 Search Gen Search Eligible for branded queries; no exclusions
04 Shopping 1 Shopping Undefined role, eligible for PMax coverage
05 Shopping 2 Shopping Same product coverage as Shopping 1 + 3
06 Shopping 3 Shopping Third Shopping with no query boundary
07 PMax 1 Perf. Max No Customer Match; competing with Shopping
08 PMax 2 Perf. Max Duplicate PMax, no audience signal
09 PMax 3 Perf. Max Third PMax; conversion volume fragmented
10 DSA Dynamic Search Overlapping with existing Search campaigns
11 YouTube Video No defined role relative to other campaigns
12 Display Display Unmeasured impact; no placement controls
13+ 9 additional campaigns with no defined roles documented
2.59× prior blended ROAS  ·  Jan–Oct 2025
After  ·  Campaign Architecture Specification
Tier Campaign Type Role Budget Priority Key Exclusions
Tier 01 Brand Defense Search Capture brand-name searches; protect branded demand from generic campaigns Secondary Non-brand terms excluded via negative list
Tier 02 Core Shopping Shopping Structured product groups covering high-intent model-specific queries Primary Brand terms and PMax overlap excluded
Tier 03 Performance Max Perf. Max Broader discovery anchored by Customer Match purchaser signals Supporting Brand exclusion applied; Shopping query overlap managed via campaign priority
4.14× blended ROAS  ·  Nov 2025–May 2026

The order mattered: feed first, bidding last

Each layer depended on the one before it working correctly. Starting with campaign structure before the feed was accurate would have produced structurally clean campaigns pointing at the wrong queries.

Rewrote every product title to match what buyers actually search for

Titles rebuilt to lead with category, compatibility, and brand: "iPhone 16 Case Clear MagSafe Compatible" instead of "SKU-AJ7729-IP16-CLR." Impression share on high-intent queries moved within the first week.

Consolidated 20+ campaigns around three defined roles

Brand Defense, Core Shopping, Performance Max: each with a clear job and defined query ownership. Individual campaigns were structured within those roles. The logic changed, not just the count.

Built negative keyword lists that kept each campaign in its lane

Negative keyword lists for each tier enforced the boundaries. Each campaign owns its query type. Budget flows accordingly.

Connected the client's customer list to PMax as audience signals

The CRM had never been linked to the ad account. We uploaded verified purchaser history as Customer Match signals. Performance Max finally had a real reference point for who it was looking for.

Reset bid targets to what the account had actually been doing

We pulled 90-day conversion baselines and set targets against those. Once realistic, the bidding system started spending. Conversions followed.

The feed is the first thing we look at in any Google Shopping account. Most accounts we take over have had years of bid tuning on top of a feed that was never fixed.

Campaign Architecture

Campaign architecture: three defined tiers

Campaign Type Query Ownership & Role Budget Priority
01
Brand Defense Search
Brand-name searches using exact and phrase match. Competitor terms excluded. Negative keyword walls prevent crossover into Tier 02.
Secondary
02
Core Shopping Shopping
High-intent category and compatibility queries. Feed titles rebuilt to match search behavior: model, finish, compatibility. Primary budget holder.
Primary
03
Performance Max Perf. Max
Feed-led discovery and expansion. CRM purchaser list uploaded as Customer Match signals. Brand and query controls reduce overlap with Search tiers.
Supporting

The three tiers describe campaign roles, not raw campaign count. Within each tier, individual campaigns carry a defined query boundary and do not overlap with the others. Every campaign now has an exclusive territory.

Seven months later: 4.14× blended ROAS

Before
20+ campaigns with overlapping query coverage across all tiers
Feed titles built from SKU codes and internal model references
Bid targets set to aspirational numbers, not historical performance
After
Three-tier structure: brand, category, and catchall with defined query ownership
Feed rewritten to match search behavior: model, finish, compatibility
Targets calibrated from 90-day conversion baselines, enabling the system to spend

$6.6M tracked revenue  ·  Nov 2025 – May 2026 (7-month post-rebuild)  ·  $1.6M managed spend

The feed rewrite changed which queries the products were eligible for, and that determined everything downstream: campaign structure, bid efficiency, and eventually revenue. The structural changes had no meaningful effect on spend; the improvement came entirely from where that spend went.

Phil Chan Head of Growth MESAscale
View on LinkedIn
About the strategist

Phil brings more than a decade of paid-media experience across agency, enterprise, and ecommerce environments. Most of that time was spent in senior account leadership, directing multi-channel portfolios at seven-figure monthly scale.

He now works directly with established ecommerce brands through MESAscale, focusing on Google, Meta, and YouTube. His work centres on account architecture: where spend is being lost, how cross-channel coordination is failing, and which structural changes compound over time.

10+ Years in Paid Media Enterprise + Ecommerce Google · Meta · YouTube
Free Ecommerce Growth Diagnostic

Find out whether your feed is limiting what bidding can do.