Overview
How many people came to your site, where they dropped out, and what a visit was worth.
Everything here is measured by GA4 on your website. It counts visits and events, not orders, so it will not match Shopify line for line — and it is not meant to.
The cards
Read the top row and you know which half of the problem you have — too few people arriving, or too few of them buying.
SessionsSessionsTraffic to your listings; only lifts sales if conversion holds. · Total UsersTotal UsersCompare with sessions to gauge how often people return. · Total RevenueTotal RevenueCross-check against Shopify to confirm tracking is firing. · Revenue Per UserRevenue Per UserWhat an average visitor is worth to you. · Purchase To View RatePurchase To View RateHow convincing your product pages are to browsers.4% or more is healthy · Conv. RateConv. RateWebsite visits that ended in a purchase. Not the same as Google Ads Conv. Rate, which divides by clicks.3.5% or more is healthy · Avg. Purchase RevenueAvg. Purchase RevenueYour typical order size as GA4 sees it. · TransactionsTransactionsPair with sessions to sanity-check conversion and spot tracking gaps. · Average Session DurationAverage Session DurationA rough read on how engaged your visitors are. · Bounce RateBounce RateA quick signal of landing-page relevance and speed.Under 25% is healthy
Each carries the value, the change against the previous period, and a trend line.
Five of them — Revenue Per User, Purchase To View Rate, Conv. Rate, Avg. Purchase Revenue and Bounce Rate — also carry a graded badge. The counts do not, because a session or order count is not good or bad on its own.
Read them in three moves
- Did traffic move? Sessions · Total Users
- Did it convert? Conv. Rate · Transactions
- Was it worth anything? Total Revenue · Revenue Per User
Bounce Rate and Average Session Duration explain a bad step 2. Read them only once you know there is one.
Two pairs worth checking
Sessions + Total Users tells you whether growth is new people or the same people coming back more often. Sessions climbing while users sit flat is repeat visiting, and it usually needs a different fix than a traffic dip.
Conv. Rate + Purchase To View Rate splits the site from the product page. Conv. Rate counts every session; Purchase To View Rate counts only people who looked at an item. If the second is healthy and the first is weak, visitors are landing somewhere that never shows them a product.
The charts
The cards tell you that conversion moved. These tell you where in the buying journey it moved, and on which days.
Purchase Funnel
Where the visits go and where they stop: Page View → View Item → Add to Cart → Begin Checkout → Purchase.
Hover any stage. The tooltip gives the session count, the share that dropped out there, the share that carried on to the next stage, and the share that made it all the way to Purchase.
Read it in three moves
- Find the biggest single drop. Not the lowest stage — the widest step between two stages
- Name what that step depends on. Page View → View Item is navigation and search. View Item → Add to Cart is the product page, price and trust. Add to Cart → Begin Checkout is shipping cost and cart friction. Begin Checkout → Purchase is payment options and form length
- Change one thing there and watch the same step next period. A funnel with two changes in it explains nothing
Conversion Rate, Total Revenue & Avg. Purchase Revenue
Shows: Conversion Rate as a line on the percentage axis, Total Revenue as bars in your currency, and Avg. Purchase Revenue as a third line.
Avg. Purchase Revenue has no visible axis of its own. Read its shape against the others and get the number from the tooltip.
Read it in three moves
- Start with the bars. Total Revenue is the outcome — everything else is an explanation of it
- Then the conversion line. Bars up with the line flat means more traffic did the same thing. Bars up with the line rising means the same traffic did better
- Then the third line. Revenue up while Avg. Purchase Revenue falls means more, smaller orders — a discount or a cheaper product mix, not a better site
The last day in range is partial and will always look like a collapse. Ignore the right-hand edge.
The tables
One of these hands you a market to fix. The other hands you a day to explain.
Country Overview
One row = one country.
Columns: Country · Sessions · Conversion Rate · Avg. Purchase Revenue · Revenue Per User, each followed by a % △ against the previous period. Opens sorted by Sessions, highest first.
The header here reads Conversion Rate while the card above says Conv. Rate. Same measure, two labels.
How to scan it
- Sort by Sessions, highest first
- Look across at Conversion Rate
- The row that matters a country in your top five for traffic with a conversion rate well under the others — usually currency, shipping cost or a language the page does not speak
Then sort by Revenue Per User and read Sessions beside it. A country worth a lot per visitor and sending very few of them is the one to buy more traffic in.
Date Overview
One row = one date, oldest first.
Columns: Date · Total Revenue · Purchase To View Rate · Cart To View RateCart To View RateSee the full entry. · Bounce Rate.
This table has no % △ columns and no Compare option — it is a run of days, and the day before it is already the row above.
How to scan it
- Sort by Total Revenue, highest first, to find your best days
- Look across at Cart To View Rate and Purchase To View Rate on those rows
- The row that matters a day with strong Cart To View Rate and weak Purchase To View Rate — people wanted the product and something after the cart stopped them
Bounce Rate is the fastest way to spot a day when a campaign sent the wrong traffic. It moves before revenue does.
Filters
Cutting this page to one traffic source is the fastest way to tell a campaign problem from a site problem.
Source and Medium are multi-select dropdowns, both defaulting to All. Pick your values and press Apply — nothing refetches until you do. They re-filter the cards, the funnel, the chart and both tables together.
The property switch appears only when more than one GA4 property is connected. Switching it clears your Source and Medium picks back to All and reloads the page, because those values belong to the old property.
The date range and country filter at the top of the app also apply here.
Search inside a table matches the first column and filters rows only. The summary row is hidden while a search is active, because it totals every row in range.
Each table has a view switcher to bar, line or pie. Country Overview opens on Sessions, Date Overview on Total Revenue.
Watch out for
They are counted differently and will not agree. GA4 only records a purchase when the tracking fires in the browser, so blocked scripts, ad blockers and abandoned redirects go missing. Use this number to check that tracking is alive, and Shopify for what you actually sold.
Check Total Users on the same range. Sessions down with users flat means people visited less often, which is a retention story, not an acquisition one.
It moves with traffic mix as much as with the site. A new source sending less relevant visitors raises it without a single page changing.
Confirm you pressed Apply. The dropdown holds your selection until you do, and the page keeps showing the old filter until then.
Do this first
Open the funnel and find the widest step. That one step is where the largest number of interested people are being lost, and every card on this page is a consequence of it.
See yoursYour GA4 funnel, cards and country split for the last 30 days.
Open Overview →