E-Commerce Metrics That Matter

A store can drown in charts and still not know if last week made money. Metrics that matter are the few numbers that describe the operating loop: traffic quality, conversion, basket, contribution, acquisition cost, and whether parcels actually arrived.

The longer reporting map is the e-commerce analytics guide. This article is the small dashboard and the vanity list to ignore.

A small operating dashboard

Use one table, one week vs prior week vs prior year if you have it. Definitions locked in a footnote.

Metric	Formula / source	Why it is on the board
Revenue (defined)	Commerce + finance	The top of the P&L you mean
Orders	Count of paid / delivered (say which)	Volume
Conversion rate	Orders / sessions	Improve CVR
AOV	Revenue / orders	AOV explained
Contribution $ or %	After COGS, fees, variable ship, returns estimate	Survival; margin
Ad spend	Media invoices	Input
MER or blended ROAS	Revenue / ad spend	Directional efficiency; ROAS
Paid CAC	Paid spend / new customers	CAC
Fulfillment health	On-time %, RTO, ticket volume	Ops; shipping

Calculators for the ratios: conversion rate, AOV, ROAS.

Monthly (not weekly noise): LTV by cohort, return rate by SKU, contribution after ads vs overhead.

Channel cuts: paid vs organic vs email vs marketplace—same definitions, not five different “revenues.”

Vanity metrics to ignore (or demote)

• Raw sessions without conversion and contribution. Bots and brand spikes lie.
• Rankings as a KPI. Rank without revenue is a hobby. Use Search Console as diagnosis, not as a trophy.
• Email list size without revenue per send and unsubscribe rate.
• Social followers.
• Pageviews per session as a goal (can mean confusion).
• Platform ROAS pasted as profit (ROAS vs ROI).
• Add-to-cart rate in isolation (can rise while checkout dies).
• Time on site.

You may still look at these when diagnosing. They should not be the scoreboard in a Monday meeting.

Identity: revenue and orders

Write down:

• Gross merchandise vs net of discounts.
• Whether shipping charged to the customer is in revenue.
• Prepaid vs COD: placed vs delivered.
• Time zone and whether refunds reverse the week they happen or the week of the order.

If marketing uses checkout GMV and finance uses recognized revenue, both can be right—and the dashboard needs two rows or one agreed row.

How the metrics talk to each other

Revenue ≈ sessions × CVR × AOV (approximately). If revenue is down, you know which term moved.

Contribution is the filter: a CVR win from a 40% off sitewide sale can be a contribution loss.

CAC vs first-order contribution is the acquisition filter. LTV is the “we can wait” filter—use ranges.

Payment success belongs next to CVR: payment failures can look like a merchandising problem.

Cadence and owners

• Daily: site up, checkout up, gateway settlement vs orders (exception watch).
• Weekly: the table above, plus top SKUs and top leaks (a landing page, a decline code, a courier).
• Monthly: cohorts, inventory turns, app/software cost vs volume.

Assign a human to each weekly number. Unowned metrics do not change.

What “good” looks like without fake benchmarks

Do not pin a public “average conversion rate” on the wall. Mix, price, and COD vs prepaid change the number. Good is:

• You can explain a week’s revenue move with CVR, AOV, or sessions—not with a story.
• Contribution after ads is a signed number you trust, even if it is small.
• Fulfillment exceptions are visible before they become chargebacks.
• New-customer CAC is next to first-order contribution, not next to GMV.

If a metric cannot change a decision this month (pause a campaign, delist a SKU, staff a shift), it does not belong on the weekly board. Park it in a diagnostic tab.

SKU and landing cuts (still a short list)

Once the storewide table is stable, add two slices only:

• Top 20 SKUs by contribution (not by revenue). A hero SKU with 8% contribution after returns is a different conversation than a quiet SKU at 45%.
• Top landing pages with CVR and revenue. That connects SEO and ads to conversion work without making “rankings” a KPI.

Marketplace vs D2C should be separate rows if fees differ. Blending Amazon GMV with site GMV without fee-adjusted contribution is how business models get confused in a spreadsheet.

Tooling without a second source of truth

Commerce platform + ads UI + a sheet is enough to start. When you add analytics:

• One session definition.
• One currency.
• UTMs that do not create duplicate landing URLs (technical SEO cares; so does attribution).

If two tools disagree on orders, pick finance as the tie-breaker for money and the commerce platform for operations. Do not average them.

Seasonality: compare to the same week last year when you have it, not only to last week. A “down week” after a festival sale is not a conversion emergency.

Build the dashboard in the tools you already have. Buying another analytics suite does not pick the metrics. The analytics guide is the implementation path. Watch the P&L through a short list. Everything else is supporting evidence.

The Unified E-Commerce Metric Hierarchy

Store operators are frequently overwhelmed by conflicting metrics across ad accounts, analytics platforms, and accounting software. High-growth enterprises organize metrics into three operational tiers:

Level 1: Executive Solvency & Enterprise Health
└── Contribution Margin ($) | Net Cash Flow | Marketing Efficiency Ratio (MER)

Level 2: Unit Economics & Capital Efficiency
└── Customer Acquisition Cost (CAC) | Customer Lifetime Value (LTV) | Average Order Value (AOV)

Level 3: Funnel & Tactical Performance Diagnostics
└── Storefront Conversion Rate | ROAS by Ad Creative | Return / RTO Rate | Cart Abandonment %

Cohort Repurchase Modeling

True enterprise value in digital commerce is created through customer retention. Construct monthly cohort retention curves tracking what percentage of customers acquired in a specific month make a second, third, or fourth purchase over the subsequent 12 months. If your 90-day repurchase rate exceeds 25%, you can comfortably invest more capital into customer acquisition than competitors who rely purely on single transactions.

Establishing a Weekly KPI Cadence

Set up a Monday morning operating rhythm reviewing: 1) Gross sales vs. net sales after returns; 2) Blended MER (Total Revenue ÷ Total Marketing Spend); 3) Carrier on-time delivery rates; and 4) Top-selling SKU stock availability to detect inventory replenishment risks before stockouts occur.