Seven-axis shopping benchmark constellationBlue shape is current attainment; orange ring is target
Marketing & Advertising
Shopping Campaign Benchmark Calculator
Diagnose a shopping campaign as a seven-control commerce system instead of ranking one headline metric. The benchmark respects whether higher or lower is beneficial, uses entered targets applicable to the assortment and auction context, and exposes feed disapproval, availability, response, and contribution gaps beside the composite score.
Input evidence: compare one market, product set, device mix, and recent reporting window. Enter feed rejection from diagnostics, approved conversion from commerce records, contribution ROAS from finance, and document the source of every target or ceiling.
Weighted commerce health—
Benchmark band—
Largest normalized gap—
Controls meeting gate—
Feed readiness score—
Commerce efficiency score—
Direction-aware control constellation
Compare eligibility, demand capture, economics, and availability without rewarding the wrong direction
Lower-is-better controls are inverted before normalization; the target ring is 100% attainment.
| Control | Current | Entered gate | Desired direction | Weight | Attainment | Reading |
|---|
How to use the shopping benchmark
Freeze definitions before comparing current controls with targets
- Choose one market, product scope, attribution window, and reporting period.
- Enter eligible impression share and in-stock click share from platform and catalog diagnostics.
- Enter disapproval and CPC as lower-is-better controls with explicit ceilings.
- Use approved conversion and contribution ROAS from reconciled backend orders.
- Set targets from an applicable operating standard, controlled test, or approved plan—not an unrelated industry average.
- Read the weakest normalized gap and the raw control register before using the composite score.
Shopping benchmark fundamentals
Seven controls describe different failure modes
Eligibility
Share of auction opportunities where products can serve.
Feed validity
Absence of disapproval or policy-blocking product errors.
Demand capture
Click response after an eligible impression.
Auction price
Media cost required for each purchased click.
Approved conversion
Backend accepted orders per purchased click.
Contribution ROAS
Post-margin contribution divided by media spend.
Availability
Click share landing on sellable inventory.
Result interpretation
Separate overall health from the control that needs action
Commerce health
Direction-aware weighted attainment across all seven controls.
Benchmark band
A descriptive planning label for the current composite, not an industry percentile.
Largest gap
The control with the greatest weighted shortfall after direction is normalized.
Controls at gate
Count meeting their own target or ceiling; it prevents a composite from hiding misses.
Feed readiness
Combined reading of eligibility and in-stock coverage.
Commerce efficiency
Combined conversion and contribution quality, distinct from traffic delivery.
Calculation method
Normalize every control in its beneficial direction
Higher-is-better attainment equals actual ÷ target. Lower-is-better attainment equals ceiling ÷ actual. A weighted geometric mean limits compensation when one essential control is weak.
Target selection
An applicable target must match assortment and auction context
Brand terms, generic categories, seasonal peaks, clearance products, and new feeds can have materially different CPC and conversion behavior. Record the target source and review date.
Composite interpretation
The score prioritizes diagnosis rather than league tables
A score near 100 means the entered weighted gates are collectively met. It does not prove causal incrementality, profitability across every SKU, or superiority to another advertiser.
Conflict resolution
Efficiency and availability can point in opposite directions
High contribution ROAS can coexist with low impression share because delivery is constrained. High impression share can coexist with weak economics because the campaign buys poor demand. Use the raw register to identify the operating response.
How to read the visualization
Read the constellation as imbalance, not decorative symmetry
- Meaning and scale
- Each spoke is one control normalized to its own gate; the target ring equals full attainment even though the raw units differ.
- Inputs that move it
- Changing a current value moves its blue vertex; changing a target or ceiling changes the normalization and can move the same vertex without operational change.
- Decision pattern
- Deep inward notches identify weak controls, while the table preserves raw values, direction, weight, and the action priority.
- Misleading boundary
- Area and visual symmetry are not statistical scores. A large polygon can still conceal one commercially binding control, and targets from incomparable markets invalidate the shape.
Detailed calculation process
Convert raw controls into a direction-aware weighted score
1. Higher-is-better attainmentRᵢ = Aᵢ ÷ Tᵢ
2. Lower-is-better attainmentRᵢ = Tᵢ ÷ Aᵢ
3. Weighted geometric scoreS = 100 × exp[Σ wᵢ ln(clamp(Rᵢ, 0.05, 1.25)) ÷ Σwᵢ]
4. Priority gap for each controlGᵢ = wᵢ × [1 − min(Rᵢ, 1)]Higher Gᵢ identifies a larger importance-adjusted shortfall.
In plain language: put every control into a common beneficial-direction attainment ratio, apply its decision weight, limit excessive overperformance from compensating for a failure elsewhere, and rank the weighted gaps that remain.
Percent current and target values may remain in percentage-point form because their ratio is dimensionless. CPC uses currency/click on both sides; contribution ROAS is dimensionless.
- Aᵢ
- current value for control i; control-specific unit
- Tᵢ
- target or ceiling for control i; same unit as Aᵢ
- Rᵢ
- direction-adjusted attainment; dimensionless
- wᵢ
- importance weight; percent of composite
- S
- weighted benchmark score; index points
- clamp
- scoring guardrail from 0.05 to 1.25; raw values remain in the table
Default substitution
Eligible impression attainment = 62 ÷ 70 = 0.8857. Feed disapproval attainment = 2.0 ÷ 4.2 = 0.4762 because lower is better. CTR attainment = 1.8 ÷ 2.1 = 0.8571.
CPC attainment = 1.10 ÷ 1.24 = 0.8871. Approved conversion = 3.1 ÷ 3.6 = 0.8611. Contribution ROAS = 1.62 ÷ 1.85 = 0.8757. In-stock click share = 93 ÷ 97 = 0.9588.
Each ratio enters the geometric mean with weights 16, 12, 12, 12, 18, 20, and 10. The live score is the exponentiated weighted log average.
Reconciliation: the seven weights sum to 100%; recomputing the weighted log terms from the ledger reproduces the headline score. The lowest ratio identifies feed disapproval as the largest normalized gap under defaults.
Evidence discipline
Keep platform diagnostics and commerce outcomes on compatible scopes
- Freeze market, device, campaign, and product inclusion rules.
- Use the same currency and tax treatment for CPC and contribution.
- Reconcile approval and returns after the relevant maturation window.
- Record target owner, evidence source, and expiration date.
Model limitations
The benchmark is an entered control system, not an industry truth
It excludes statistical uncertainty, SKU distributions, bid strategy dynamics, inventory replenishment, seasonality adjustments, marginal response, attribution error, and causal lift. Weight choices materially change the composite.
Key terminology
Shopping benchmark glossary
- Eligible impression share
- Received impressions relative to eligible opportunities.
- Feed disapproval
- Share of submitted products blocked from serving.
- Approved conversion
- Accepted orders per purchased click.
- In-stock click share
- Clicks landing on products available to sell.
- Contribution ROAS
- Contribution after product and fulfillment cost divided by media.
- Directional gate
- A target evaluated as higher or lower is better.
- Geometric mean
- Composite that limits compensation across weak dimensions.
Practical decision cases
Use the weak control to choose the next intervention
Healthy traffic, broken feed
CTR and CPC meet gate, but disapproval and in-stock share create the deepest notch. The next budget goes to diagnostics and inventory coverage rather than bid expansion.
Efficient orders, limited reach
Approved conversion and contribution ROAS are strong while impression share is low. The team tests auction headroom only after confirming the current contribution survives added scale.
All gates nearly met
No single metric is severely weak, but the weighted score remains below the approval threshold. The decision is to verify target provenance and sampling stability before launching a broad remediation plan.
Important note
Before relying on this result
The benchmark excludes sampling uncertainty, seasonality normalization, marginal bid response, SKU distributions, attribution error, replenishment, and causal effect. Targets and weights are user-governed.
Additional Shopping Campaign Benchmark Calculator questions
Can industry averages be entered as targets?
Only when their product mix, market, period, definitions, and measurement quality are sufficiently comparable; an internal operating gate is often more useful.
Why is feed disapproval inverted?
Lower disapproval is better, so attainment equals the entered ceiling divided by the current disapproval rate.
Can a strong metric compensate for a failed control?
Only partially. The geometric composite limits compensation, and the raw register still identifies every failed gate.
Is the score a prediction of sales growth?
No. It is a direction-aware control index under entered weights and targets.