Research Study
The Account Structure Profit Study
Why profit-based SKU clustering outperforms category structures in Google Ads
Key Finding
Profit-tier SKU clustering consistently delivers higher profit efficiency and faster recovery from disruption than generic category-based structures, particularly in multi-SKU ecommerce accounts.
Abstract
This study examines whether Google Ads account structure still influences commercial outcomes in an era of automated bidding. Specifically, it compares traditional category-based structures with profit-based SKU clustering to assess their impact on profit efficiency, budget allocation, and resilience to operational change.
Background
A common belief in paid search is that account structure no longer matters. With smart bidding and Performance Max, many teams assume Google's algorithms will self-correct regardless of how products are grouped.
This study tests that assumption using real ecommerce account data, focusing not on revenue or ROAS, but on profit and budget efficiency.
Dataset
Accounts analysed
12 DTC ecommerce brands
Monthly spend range
£8,000 to £120,000
Product catalogue size
150 to 6,000 SKUs
Observation period
12 to 16 weeks per account
Channels
Google Shopping & Performance Max
Primary metric
Gross profit & POAS
All data was anonymised. Accounts were observed across two different structural approaches at separate time periods. Seasonality and major pricing changes were controlled for where possible.
Structures Compared
Structure A
Category-Based (Standard)
Campaigns or asset groups organised by:
- • Product category
- • Collection
- • Brand line
Shared budgets across SKUs with mixed margins, velocity, and profit contribution.
This structure reflects common "best practice" setups used by many agencies and in-house teams.
Structure B
Profit-Tier SKU Clustering
Products grouped by profit density, not taxonomy:
- Tier 1: High margin, high velocity
- Tier 2: High margin, low velocity
- Tier 3: Low margin, high velocity
- Tier 4: Long-tail and exploratory SKUs
Each tier operates with independent budgets, tier-specific scaling rules, and different tolerance for volatility and experimentation.
Methodology
For each structure, the following were measured weekly:
- • Profit per £1 spent (POAS)
- • Alignment between spend share and profit share
- • Budget reallocation efficiency during disruption events
- • Time-to-recover to baseline profit after disruption
Disruption events included: stockouts, promotion start and end, price changes, and demand shifts. Results were aggregated across accounts and normalised to account size.
Results
1. Profit Efficiency
Profit-tier SKU clustering delivered:
12-22% higher POAS
compared to category-based structures
Strongest gains observed in catalogs exceeding 500 SKUs and brands with margin variance above 15 percentage points.
2. Budget Allocation Accuracy
Category-based structures consistently allocated 25-40% of spend to below-average profit SKUs.
Profit-tier structures aligned spend much more closely with profit contribution, reducing inefficient budget leakage.
3. Volatility and Recovery
During disruption events:
- • Category structures showed slower recovery and higher spend inertia
- • Profit-tier structures recovered 30-45% faster to baseline profit levels
Budgets rebalanced cleanly between tiers without requiring frequent manual intervention.
4. Scaling Behaviour
Category structures increased revenue but often flattened or reduced profit as spend scaled.
Profit-tier structures preserved POAS until clear, tier-specific saturation points were reached. This allowed controlled scaling rather than blanket budget increases.
Interpretation
Automated bidding systems optimise within the constraints of the structure provided.
When SKUs with different profit characteristics are grouped together, the algorithm optimises toward volume and blended efficiency, not profit.
Profit-tier clustering sharpens the signal by aligning budget boundaries with economic reality.
Account structure is not obsolete.
Poor structure is simply masked by automation.
Practical Framework
Profit-tier clustering recommended when:
- • Product catalogue exceeds 200 SKUs
- • Margin variance exceeds 15 percentage points
- • Monthly Google Ads spend exceeds £5,000
Simpler structures may be sufficient when:
- • Catalogs are small
- • Margins are flat
- • Spend is tightly constrained
Limitations
- • Observational study, not a simultaneous A/B split
- • Profit attribution dependent on data quality
- • Findings strongest in mid-to-large ecommerce accounts
These limitations do not undermine the directional consistency of the results.
Conclusion
Profit-based SKU clustering materially improves profit efficiency, budget control, and resilience in Google Ads accounts.
Category-based structures optimise for organisational neatness.
Profit-tier structures optimise for money.
For ecommerce brands scaling beyond small budgets, the distinction matters.
Stay informed
Get our latest insights on Google Ads strategy delivered to your inbox. No fluff, no spam, just honest thinking.
Unsubscribe anytime. No spam, ever.