Ecommerce guide · 10 Nov 2026 · 6 min read
Ecommerce analytics setup checklist: what to have working before you launch
An online store that cannot tell you where orders come from is guessing. This checklist walks through the set-up in the order that avoids the usual gaps, from account access to the first report.
The short answer
Set up ecommerce analytics in this order: create the analytics property and keep ownership with the business, add the tracking tag, send the standard shopping events from product view to purchase, test each one, handle cookie consent, filter out your own visits, mark purchases as a key event, and build one simple report you will actually read.
Start with ownership and a written question list
Before any tag is added, decide who owns the account. The analytics property, the tag manager container and the ad accounts should sit under a login that belongs to the business, with agencies and freelancers added as users. If a developer creates everything under a personal email, you may lose your history the day that relationship ends.
Then write down five or six questions the store must answer. For example: which channels bring orders, which products are viewed but rarely bought, and at which step of the checkout people leave. Every item on the rest of this checklist exists to answer one of those questions. If a tracking idea does not map to a question, leave it for later.
Create the property and install the tag
Most stores use Google Analytics 4 (GA4). Create a property, add a web data stream for your store's address, and note the measurement ID. There are three common ways to install it, and you should pick one and stay with it so the same event is never sent twice.
- The official integration or app for your store platform, if it has one.
- Google Tag Manager, which keeps all tags in one place. The guide to Google Tag Manager basics explains the container idea.
- Code placed directly in the theme, which is the hardest to maintain.
A common fault is two methods running at once, for example a plugin plus a manually pasted tag. That doubles page views and can double purchases. Check once the tag is live that each page view is counted a single time.
The shopping events to send
GA4 has a set of recommended ecommerce events. These names are what its built-in shopping reports expect, so use them exactly rather than inventing your own. Check the current documentation for the full parameter list for each one.
| Event | When it fires | What it tells you |
|---|---|---|
| view_item_list | A category or search results list is shown | Which lists get seen |
| view_item | A product page is opened | Interest in a product |
| add_to_cart | An item goes into the basket | Intent to buy |
| begin_checkout | The customer starts checkout | Where the funnel begins |
| add_payment_info | Payment details step is completed | How many reach payment |
| purchase | The order is confirmed | Orders and revenue |
If you offer a wishlist, an add_to_wishlist event is also available; see wishlists and save for later for why that signal is useful. Each event should carry the product identifier, name, quantity and price, and the purchase event needs a unique transaction ID so that a refreshed confirmation page is not counted as a second order.
Test before you trust the numbers
Place a real test order and watch the events arrive. GA4 has a debug view and Tag Manager has a preview mode. Both show the events as they are sent, with their parameters.
- Does each event fire once, and only once?
- Does the purchase value match the order total? Decide whether your figure includes tax and shipping and keep that choice consistent.
- Do the product names and identifiers match what is in your catalogue?
- Does the purchase event still fire when the customer pays through an external payment page or wallet and returns afterwards?
- Does it work on a phone as well as a desktop?
Refund the test order afterwards and mark it clearly in your records so it is not mistaken for a sale. For a fuller explanation of how events become goals, read GA4 events and conversions explained.
Consent, privacy and your own traffic
Rules on cookies and tracking differ between the UK, the EU, Australia, Singapore, Japan, the UAE and the different US states. Ask a qualified adviser what applies to the places you sell to. Your cookie banner should only fire analytics tags in line with the choice the visitor made, and the privacy policy should name the tools you use.
Google offers a consent mode that adjusts how tags behave depending on the visitor's choice. Check the current documentation to see how it fits your set-up. Expect some purchases to be missing from analytics because some visitors decline tracking, so compare analytics totals with your order system occasionally rather than assuming they match.
Also stop your own visits from polluting the data. Define internal traffic by IP address and apply the filter, and do not place staff test orders through the live checkout without marking them.
Mark what matters and connect other tools
In GA4, a conversion is now called a key event. Mark purchase as one. For a store that also relies on enquiries, such as a wholesale form, mark the form submission as well. Keep the list short; if everything is a key event, nothing stands out. Conversion tracking explained covers the idea in more detail.
If you run Google Ads or Meta ads, link the accounts and set up their conversion tracking from the same purchase event, so the ad platforms and analytics agree on what counts. Link Google Search Console too, so you can see which search terms lead to product pages.
Finally, think about order data that never passes through the browser, such as phone orders or marketplace sales. These will not appear in the site's analytics, so note them separately in your reports.
Common set-up mistakes to avoid
A few faults appear again and again. Knowing them in advance saves weeks of unreliable reports.
- Tracking the wrong page. Counting a visit to the thank-you page as a purchase also counts every refresh and every bookmark. Fire the event from the order itself, with a unique transaction ID.
- Mixed currencies. If you sell in several currencies, send the currency code with every value, or revenue will be added up as if it were all one currency.
- Lost referrals. When shoppers leave for an outside payment page and come back, the return can be recorded as a new visit from the payment provider. Check the list of referral exclusions in your analytics settings.
- Different names for one product. If a product has one name on the page and another in the basket, reports split it in two.
- No record of changes. Keep a short log of every tag or setting change, with the date. When a number jumps, the log is the first place to look.
None of these is hard to fix once found, but each can quietly distort decisions about advertising and stock for months.
Build one report and a routine
Set up a single dashboard that answers your original questions. A good start is sessions by channel, orders, conversion rate, average order value and the funnel from product view to purchase. The measures are defined in ecommerce metrics: conversion rate, AOV and more.
Look at it weekly, write one line about what changed, and act on at most one thing at a time. When the funnel shows a large drop between two steps, the guide on why shoppers abandon checkout helps you decide what to examine first.
Re-run the test order after any change to the theme, checkout, payment provider or cookie banner, because these are the usual causes of silently broken tracking. If you would like help with set-up on a new or existing shop, Liftup Technologies offers ecommerce solutions that include tracking planning.
Related guides
- Ecommerce solutions
- Why Shoppers Abandon Checkout and What to Fix
- Wishlists and Save for Later for Online Stores
- Ecommerce product page checklist
- Ecommerce website vs mobile app for your store
- How to start an online store steps in order
Questions and answers
Which analytics tool should an online store use?
Most stores start with Google Analytics 4 because it is free to begin with and has built-in shopping reports. Your platform may also include its own reports. Whichever you pick, use one install method and check the figures against your order system.
Which ecommerce events matter most?
View item, add to cart, begin checkout, add payment info and purchase give you a basic funnel. Add a wishlist or search event later if those features exist. Use the standard event names so the built-in reports work.
Why do my analytics orders differ from my shop orders?
Common causes are visitors who decline tracking, ad blockers, a failed purchase event after an external payment page, test orders and refunds. Small differences are normal; large or growing gaps mean something is broken.
Do I need a cookie banner for analytics?
Many places require consent or at least notice before analytics cookies are set, but the rules vary by country and state. Ask a qualified adviser for the markets you sell to, and make sure your tags respect the visitor's choice.
How do I stop counting my own visits?
Define internal traffic by IP address in your analytics settings and activate the filter. Also avoid placing unmarked test orders on the live store, and record any you do place so they can be excluded from sales reports.
How often should I check store analytics?
A short weekly look at channels, orders and the checkout funnel is enough for most stores. Check more closely after any change to the theme, checkout or payment set-up.
Tell us what you are building.
A short call, a written plan and a fixed quote. No obligation.