The Marketplace KPI Glossary
Definitions, calculations, and tactical interpretation
KPIs are only useful if the whole team is using the same math to make decisions. This glossary covers the core metrics of marketplace operations, how to calculate them, and—more importantly—what they're trying to tell you.
1) Traffic & advertising efficiency
These metrics tell you how well you're "buying" attention and whether that attention is relevant.
CTR (Click-Through Rate)
Calculation: Clicks ÷ Impressions × 100
What it tells you: Are you winning the click on the search shelf?
Operator note: CTR is usually driven by your main image, price, reviews, title, and query relevance. Low CTR typically means you're showing up for the wrong searches—or your shelf presentation isn't competitive.
Common benchmark (varies by category):
Sponsored Products: ~0.3–0.5%
Sponsored Brands: ~0.4–0.8%
CPC (Cost Per Click)
Calculation: Ad Spend ÷ Clicks
What it tells you: The cost of a "shop visit."
Operator note: High CPC is only "bad" if your conversion and margin can't support it. Premium categories often have high CPC and still print money if CVR and contribution margin are strong.
ACoS (Advertising Cost of Sale)
Calculation: Ad Spend ÷ Ad-attributed Sales × 100
What it tells you: How efficient the ad is on the sales it gets credit for.
Operator note: ACoS is great for keyword-level bid decisions, but it's not the full truth of the business because attribution is imperfect and organic halo exists.
ROAS (Return on Ad Spend)
Calculation: Ad-attributed Sales ÷ Ad Spend
What it tells you: The inverse of ACoS; "How many dollars came back per dollar spent."
Operator note: ROAS is easier for exec teams to read quickly. Just don't let it become the only metric—ROAS can look "great" while total sales stagnate.
TACoS (Total Advertising Cost of Sales)
Calculation: Ad Spend ÷ Total Sales × 100
What it tells you: Whether advertising is helping the entire business grow—or just replacing organic sales you used to get for free.
Operator note: If TACoS rises while total sales stay flat, you're often buying back demand (organic rank slipped, competition increased, or branded defense is failing).
Typical ranges:
Mature listings: ~8–12%
Launch/expansion: ~15–20%
Impression Share (where available)
Definition: Your share of eligible impressions for a keyword or target.
What it tells you: Are you actually showing up enough to win?
Operator note: If you're losing impression share due to budget, you may be "efficient" and still missing the entire opportunity.
Branded vs Non-Branded Traffic Share
Definition: How much traffic comes from brand-name searches vs generic category searches.
Why it matters:
- Branded traffic is usually cheaper and converts higher.
- Non-branded traffic is how you grow beyond existing demand.
Operator note: If non-branded is weak, you're dependent on your current audience. If branded is weak, you're losing your own demand to competitors.
Organic Sales %
Calculation: Organic Sales ÷ Total Sales × 100
What it tells you: How much of your revenue comes without paid support.
Operator note: A healthy business usually increases organic share over time as listings stabilize and rank improves.
New-to-Brand % (Amazon, Sponsored Brands / DSP contexts)
Definition: The percentage of ad-attributed customers who are new to your brand.
What it tells you: Whether ads are acquiring new customers or recycling existing demand.
2) Conversion & listing performance
Once a shopper lands on the page, these metrics track your ability to close the deal.
Sessions / Visits
Definition: Visits to the product detail page.
Operator note: Sessions are the denominator for conversion. Low sessions = noisy CVR. If you don't have enough traffic, don't overreact to week-to-week conversion swings.
CVR (Conversion Rate)
Calculation: Orders ÷ Sessions × 100
What it tells you: Does the listing close?
Tactical interpretation:
- 10–15%+ often indicates strong product-market fit and a solid listing
- 5–10% can be workable but usually has obvious upside
- <5% usually indicates a leak: price, reviews, images, variation confusion, trust issues, or mobile rendering problems
Operator note: Segment CVR where possible (branded vs non-branded, mobile vs desktop, top keywords). Blended CVR hides the real failure point.
Unit Session Percentage (Amazon)
Definition: Units ordered per session.
What it tells you: Whether customers tend to buy multiple units or bundles.
Operator note: If Unit Session % is meaningfully higher than CVR, shoppers are "stocking up," which can justify higher CPC.
Add-to-Cart Rate (where available)
Definition: The percentage of sessions that add to cart.
What it tells you: Early intent signal.
Operator note: Low add-to-cart often points to price friction or confusion (variation, sizing, unclear benefits). High add-to-cart but low purchases can indicate fulfillment/checkout friction.
Buy Box / Featured Offer % (Amazon)
Definition: The % of page views where your offer owned the Buy Box/Featured Offer.
Operator note: If this drops below ~95%, you have a system leak (pricing wars, unauthorized sellers, inventory issues). It also kneecaps ad performance.
Search Rank (Keyword Rank)
Definition: Where your product appears for a given search term.
Why it matters: Rank influences traffic quality, volume, and required ad spend.
Operator note: The most actionable rank tracking is on a set of high-intent terms, not thousands of random keywords.
Listing Health Score (internal KPI)
Definition: A structured score for the listing's readiness and quality (content, images, attributes, variations, compliance, reviews).
Why it matters: Lets you prioritize fixes across a catalog and prevent "random" performance swings.
Operator note: The goal is a weekly cadence where health scores don't drift.
3) Profitability (the bottom line)
Revenue is vanity; margin keeps the lights on.
Contribution Margin (the North Star)
Calculation:
Revenue − (COGS + Marketplace Fees + Fulfillment/Shipping + Ad Spend + Returns/Refunds)
What it tells you: What's left after variable costs.
Operator note: Many "top sellers" are unprofitable once ads and returns are included. Contribution margin prevents you from celebrating growth you're paying for.
Healthy target (varies): ~20–30%+ contribution after variable costs.
Gross Margin
Calculation: (Revenue − COGS) ÷ Revenue
What it tells you: Product margin before marketplace realities hit.
Operator note: Useful, but incomplete on marketplaces.
Net Margin
Definition: Profit after all costs (fixed + variable).
Why it matters: Executive truth metric, but less actionable daily than contribution margin.
CAC (Customer Acquisition Cost) (where measurable)
Calculation (simple): Ad Spend ÷ New Customers
What it tells you: What you paid to acquire a new customer.
Operator note: CAC is only meaningful if "new customer" is measured reliably (platform definitions vary). Pair it with LTV when possible.
LTV (Customer Lifetime Value) (where measurable)
Definition: Expected total profit from a customer over time.
Operator note: Marketplace data often limits LTV visibility. If you can't measure it, don't pretend you can—use repeat purchase rate proxies instead.
Price Index (vs competitors)
Calculation: Your price ÷ Competitor price (or your price vs category median)
What it tells you: Whether your offer is priced competitively.
Operator note: When CVR drops, price index is one of the first checks—especially if reviews are stable.
Coupon / Promo Lift
Definition: Incremental sales or CVR improvement generated by a coupon or promotion.
Operator note: Watch profitability. Promotions can increase revenue while shrinking contribution margin.
Subscribe & Save % (if applicable)
Definition: % of units/orders coming via subscription.
Why it matters: Often improves retention and stabilizes demand, but affects margin and forecasting.
Refund / Return Rate
Calculation: Returned units ÷ Units sold × 100
What it tells you: Hidden profit leak + expectation gap indicator.
Operator note: High returns often mean the listing is overselling, unclear, or the variation structure is confusing.
4) Inventory health & risk
These metrics protect you from stockouts, overstock, and cash drag.
Weeks of Cover (WOC)
Calculation: On-hand Units ÷ Average Weekly Sales
What it tells you: How long inventory will last at current velocity.
Operator note: 4–8 weeks is a common "Goldilocks" zone for many categories—but lead times change the right answer.
Sell-Through Rate
Calculation: Units Sold ÷ Available Inventory × 100
What it tells you: Are you moving inventory at a healthy pace?
Operator note: Low sell-through creates storage fee drag and dead stock. Extremely high sell-through can mean you're flirting with stockouts.
Inventory Turnover
Calculation: COGS ÷ Average Inventory Value
What it tells you: How efficiently you're using cash.
Operator note: High turnover is great—until it becomes constant stockouts.
5) Metric relationships (the diagnostic guide)
The most valuable insights come from how metrics move together.
If you see High CTR + Low CVR
It usually means: mismatch or friction.
You won the click, but the listing didn't close (price/reviews/variation confusion/weak images/unclear value).
If you see High ACoS + Low TACoS
It usually means: healthy growth.
Ads look expensive, but total sales are rising and organic is holding. You're often "seeding" rank.
If you see Low Sessions + High CVR
It usually means: untapped potential.
Your offer closes, but not enough people see it. Increase visibility through targeting, ranking strategy, or stronger shelf presentation.
If you see Flat sales + Rising TACoS
It usually means: organic erosion.
You're spending more to maintain the same result. Something is stealing your "free" sales (rank loss, competitor pressure, Buy Box loss, branded defense).
If you see Rising WOC + Falling CVR
It usually means: inventory drag.
Demand is softening or competition increased. Consider pricing, positioning, offer stack, and whether the listing needs a refresh.
Common pitfall: the blended trap
Account-wide averages hide structural problems. A 10% TACoS can look great if one hero SKU runs at 2% while five laggard SKUs sit at 40%. Always review KPIs at the SKU, category, and keyword level before making "portfolio" decisions.