Ecommerce Analytics: Which Metrics to Track, and Where the Numbers Come From

Ecommerce analytics isn't a dashboard. It's a short list of questions about your business, each answered by a metric with an exact formula and one trusted source. This guide groups the metrics that matter for a DTC brand by question, gives the definitions your tools actually use, and shows how to reconcile the numbers when your store, GA4, ad platforms and payment processor disagree. By industry: benchmarks for supplement brands, benchmarks for skincare brands, benchmarks for online pharmacies, benchmarks for telehealth companies, benchmarks for fitness brands and benchmarks for info product businesses.

What is ecommerce analytics?

Ecommerce analytics means collecting data from every system that touches an order, then using it to decide what to change. The data comes from your store platform, your analytics tool, your ad accounts, your email and SMS tools and your payment processor. None of them sees the whole picture.

The simplest way to organize it is by the question each metric answers.

Analytics guides also sort the work into four types: descriptive (what happened), diagnostic (why), predictive (what's likely next) and prescriptive (what to do). For most DTC brands, clean descriptive and diagnostic reporting comes first. A forecast built on purchase counts you can't trust is just a confident guess.

Acquisition metrics: what a new customer costs

Acquisition metrics tell you whether you can afford to keep buying growth. Shopify's own metrics guide defines them this way:16Shopify, Ecommerce metrics guide (formulas for CAC, ROAS, repurchase rate)

CAC alone says little. A $60 CAC is cheap if the first order leaves $70 of contribution margin and expensive if it leaves $20. How long it takes to earn that cost back is the CAC payback period.

Conversion funnel: where shoppers drop

The funnel tracks the same shoppers from landing to purchase. In Shopify, every step is measured in sessions, and each rate is that step's session count divided by total sessions.1Shopify Help Center, Behavior reports

Two definitions matter here:

• Shopify conversion rate is "the percentage of sessions that resulted in a purchase." Shopify warns that your order count and your sessions that completed checkout "might not be the same": one session can hold more than one purchase.1Shopify Help Center, Behavior reports
• GA4 session key event rate is the number of sessions in which a key event happened ÷ total sessions.5Google Analytics Help, Analytics dimensions and metrics If you've marked several events as key events, it's not a purchase conversion rate. Build a purchase-only rate instead.

The step between "reached checkout" and "completed checkout" mixes two very different losses: shoppers who leave, and shoppers whose payment was declined. Our guide to checkout abandonment rate shows how to separate them, and how to increase ecommerce conversion rates covers what to fix first.

Order economics: AOV, contribution margin and LTV

Order economics tells you what each sale earns. Use the platform's exact definitions, because "sales" means four different things in Shopify's data model:2Shopify Dev Docs, ShopifyQL sales schema

AOV is measured before refunds. A bundle that lifts AOV but doubles returns can look like a win on the AOV chart and a loss in net sales.

Contribution margin is what an order leaves after the costs that grow with each order: product cost, pick-and-pack and shipping, payment fees, and returns. It's the number to compare with CAC.

Customer lifetime value (LTV) extends that view over the whole relationship: average order value × purchase frequency × customer lifespan, multiplied by margin. The formulas and their traps are in what is CLTV. The full cost-side view, product by product and channel by channel, is in our guide to profitability analysis.

Retention: who comes back

Retention metrics tell you whether the first order was the start of a relationship. Shopify's metrics guide defines the basics:16Shopify, Ecommerce metrics guide (formulas for CAC, ROAS, repurchase rate)

A blended returning customer rate hides the most useful signal: whether newer customers behave better or worse than older ones. That's what cohort analysis shows, by following each month's new customers separately.

For subscription brands, add churn rate, MRR and gross vs net retention. Split churn into voluntary cancellations and failed renewals: they have different causes and different fixes, covered in subscription payment failures.

Payment metrics most dashboards miss

Searches for "ecommerce payment analytics" and "transaction analytics" point at a gap: most ecommerce dashboards stop at the purchase event. Between "clicked pay" and "money in the bank," three things can still go wrong.

How you count attempts changes the acceptance rate a lot: blocked payments, failed 3D Secure and retries all need a rule. The details are in payment acceptance rate.

The dispute rate isn't just a health metric: card networks measure it too. Visa's acquirer monitoring program (VAMP) counts fraud reports plus disputes, divided by settled card-not-present transactions, in number of transactions, not amount. Its "Excessive" merchant threshold dropped to 150 basis points (1.5%) on April 1, 2026 in Asia-Pacific, Canada, Europe and the US, with a minimum of 1,500 fraud reports and disputes a month.11Visa, Visa Acquirer Monitoring Program fact sheet, 2025 Mastercard divides one month's chargebacks by the previous month's payments.12Stripe Docs, Card network monitoring programs Track your rate the way the networks do, by count. How to keep it down is in chargeback prevention.

Inside Tagada

Build the funnel once, route every customer to the right offer

Checkout, upsells and subscriptions on one flow, with conditions per country, order or customer.

Start free

Where each number lives

Every system on your stack counts something slightly different. Knowing which one owns which metric ends most "the numbers don't match" meetings.

Why GA4 and Shopify never show the same revenue

Expect a gap. The question is whether it's stable and explainable. The usual causes, from the GA4 and Shopify documentation:

• Missed events. GA4 only counts purchases its tag received. A refused consent banner, an ad blocker or a closed tab before the thank-you page drops the order from GA4, not from Shopify.
• Duplicate events. Send each purchase once, with a transaction_id, which Google calls "required for purchases and refunds."9Google for Developers, Measure ecommerce (GA4) A thank-you page that fires on reload counts the same order twice.
• Refunds. GA4's purchase revenue is revenue "minus any refunds given," but only refunds you send as refund events. Gross purchase revenue ignores them.5Google Analytics Help, Analytics dimensions and metrics
• Currency. Set currency whenever you send value.9Google for Developers, Measure ecommerce (GA4) GA4 converts to the property's currency using the prior day's exchange rate, so multi-currency stores never match to the cent.8Google Analytics Help, Currency in Analytics
• What "value" contains. If your purchase event sends items only, while Shopify's total sales include shipping, taxes and duties, the gap is structural.2Shopify Dev Docs, ShopifyQL sales schema
• Test orders. Shopify keeps test orders and simulated transactions out of payouts and reports.4Shopify Help Center, Placing test orders Your GA4 tag can still record them as purchases. Filter internal traffic or test in a separate property.
• Time zones. Store, GA4 and processor can each run on a different time zone, so late-evening orders land on different days.

The event setup behind steps 1 and 2 is in GA4 ecommerce events. Sending the purchase from your server, once the payment has succeeded, removes most missed and duplicate events: see server-side tracking for ecommerce.

A starter ecommerce analytics dashboard

Start with one page and no more than a dozen metrics. Every line has a formula, a single source and a cadence that matches how fast you can act on it.

What to watch in real time

Most metrics are fine weekly. A few need eyes or alerts within minutes, because a broken checkout or a processor outage costs orders every hour. Know how fresh each source really is:

Shopify's Live View counts visitors who loaded a page or clicked in the last 5 minutes, in store time.15Shopify Help Center, Live View GA4's Realtime report covers the last 30 minutes, caps each card at 700 rows and has limited attribution,13Google Analytics Help, Realtime report and Google notes that data processing can take 24 to 48 hours.14Google Analytics Help, Data freshness Don't close the books on yesterday from a realtime view.

What deserves an alert: a drop in payment acceptance rate, a processor returning errors (where dynamic payment routing can send payments elsewhere), and bursts of small declined attempts that look like card testing (see ecommerce fraud prevention).

Common ecommerce analytics mistakes

The session and user definitions come from Google's documentation: a session times out after 30 minutes of inactivity by default, and session counts in standard reports are estimates.6Google Analytics Help, About Analytics sessions The attribution rules are in GA4's attribution settings: 30-day default lookback for acquisition key events, 90 days for the others.7Google Analytics Help, Select attribution settings

The tracking behind the numbers

Metrics are only as good as the events underneath. If your purchase counts don't hold up, fix collection before building more reports:

• the GA4 event list and parameters: GA4 ecommerce events and Google Analytics 4 setup;
• sessions that break across a separate checkout or payment domain: cross-domain tracking;
• purchases sent from your server, deduplicated with the browser pixel: server-side tracking;
• choosing tools for ad platforms and analytics: conversion tracking tools;
• how credit is split between channels: attribution modeling.

Whatever tools you use, the test is the same: can you put store orders, analytics purchases and processor payouts for one month side by side, and explain every gap?