Measure

Analytics

The Analytics screen shows how much revenue we can honestly connect to your cart drawer — and today it tells you, plainly, that it is not collecting yet.

What this screen is

Analytics is where Smart Cart reports what your cart drawer did: revenue it can be connected to, how many orders used it, which module was involved, and how those numbers were worked out.

Open it from the app's left-hand menu: Analytics.

Read this before you read the screen. Nothing is being collected for your store yet. The cards do not show 0 — they show a short explanation of what is missing and why. That is deliberate. A wall of zeros would look like a broken app, and would be impossible to tell apart from a real, measured zero later on. Until data arrives, treat this page as a description of what will be measured, not as a report.

Why there are no numbers yet

Two separate things feed this screen, and neither is running yet. They are reported separately because they are held up by different work.

Source	What it would feed	Status today
Storefront events (what shoppers do in the drawer)	Average order value, the funnel	Not switched on. No published cart configuration carries an analytics endpoint, so nothing is recorded.
Shopify order webhooks (what shoppers bought)	Attributed revenue, orders that used the cart, revenue by module	Not subscribed. The app does not yet request order data from Shopify, so orders cannot be matched back to a cart.

The cards say which of the two they are waiting on. "No data yet" means storefront events. "Not collecting yet" means order webhooks.

The storefront-events check is a live one, not a fixed message. The moment your store's first event arrives, the cards that depend on it stop saying "No data yet" on their own — no update or reinstall needed.

Reading the screen

Open Analytics from the app menu.
Pick a period at the top: Today, Last 7 days, Last 30 days or Last 90 days. Last 30 days is the default.
Read the sentence under the buttons. It tells you the period you are on, the period it is compared against (the same number of days immediately before), and which timezone the days are measured in.
Work down the sections: revenue, average order value, module breakdown, funnel.
Set your attribution window near the bottom if 24 hours is not right for your store.
Read How we attribute revenue at the foot of the page. It is the method, in full, not a tooltip.
Use Export CSV if you want the figures in a spreadsheet.

What each section shows

Section	What it reports
Period	The four period buttons, the comparison period, the timezone your days are measured in, and the Export CSV button.
Revenue we can account for	Attributed revenue, and the number of orders that used the cart. Followed by an "as of" note about when the figures were last correct.
Average order value	Two figures side by side: orders that used the cart drawer, and orders that did not. A permanent note beside them explains what that comparison is and is not.
By module	Attributed revenue split by which block was involved — upsell, free gift, and so on. Only appears once orders are being attributed.
Funnel	Cart opened through to checkout. Needs storefront events before it can show anything.
Attribution window	The one setting on this screen. See below.
How we attribute revenue	The published method, in five parts, plus a link to the public version of the same page.

Why the two average order values are not a before-and-after

The screen says this next to the figures and it is worth repeating here. Shoppers who open a cart drawer are already closer to buying than shoppers who never open one. So the two averages describe two different groups of people, not the same group before and after. The gap between them is a description of who bought — not a measure of what the drawer did. We will not describe it as anything stronger until there is a real A/B test behind it.

The attribution window

Shopify tells us that an order happened. It does not tell us why. So "this order is connected to the cart drawer" is a judgement, and you get to set how generous it is.

The window is how long after someone uses your cart an order still counts as connected to it.

Setting	Field	Allowed	Default
Attribution window	`Hours`, with a Save button	1 to 168 hours (one hour to one week)	24 hours

Scroll to Attribution window.
Type the number of hours you want.
Click Save. A green banner confirms the new setting.

A number below 1 or above 168 is pulled back into range rather than rejected, and anything unreadable falls back to 24. The window you set here is the same window used when orders are matched — the screen and the matching cannot drift apart.

A shorter window is stricter and will credit the cart with less. A longer one is more generous. Neither is more correct; picking one and knowing what it is beats an app that fixes the number and never tells you what it chose.

Periods, timezones and comparisons

The comparison period is always the same length as the one you picked, immediately before it — the seven days before your last seven days, not "the same week last month". Equal length is what makes two numbers comparable at all.

Days run midnight to midnight in your store's own timezone when we know it. If we have not been able to read your timezone from Shopify, the screen says so and uses UTC instead, and suggests reinstalling the app to refresh it. If you sell from Auckland and the screen says UTC, your "today" ends around midday your time — worth knowing before you read a figure.

If a date range cannot be read, you get an orange banner saying so and the screen falls back to the last 30 days. It will not quietly show you a different period and let you assume it was the one you asked for.

Exporting to a spreadsheet

Click Export CSV in the Period section. The file carries the period, the comparison period, the timezone and your attribution window in its top rows, so the numbers still mean something once the file has been renamed and forwarded.

The export is as honest as the screen. Every row has a Status column reading either collected or not collecting yet, and rows that are not being collected have an empty value cell rather than a zero. A spreadsheet cannot show "we are not measuring this" — but it also cannot add up a blank cell by accident.

The file opens correctly in Excel, including accented product names, and text cells are prefixed so a product named something like -40% Bundle is never treated as a spreadsheet formula.

How revenue is attributed

The full method is printed at the bottom of the Analytics screen, and it is also published publicly — the screen links to it — so you can send it to anyone who queries a number without giving them access to your admin. In short:

Two kinds of credit, never added together. "Direct" means an item one of our blocks put in the cart, recorded on the line at the moment it was added. "Influenced" means the shopper opened the drawer at some point before ordering. These are shown separately because they are different claims.
What a figure counts. Revenue is counted after any discounts on that line, so it reflects what you actually received. An order counts once per block, however many times that block was used. An order with a directly attributed item is not also counted as influenced.
The window. As described above — your setting, 24 hours by default.
Refunds change history, on purpose. If an order is refunded, cancelled or edited, we correct the period it belonged to rather than leaving a figure standing that is no longer true. There is no time limit on this. A report can therefore read differently today than it did last month, which is why every figure carries the time it was correct at.
What we do not claim. The average order value comparison is not a measure of effect.

Because refunds are removed from revenue but the cart was still served, an order count elsewhere in the app can be higher than the order count here. That is the refunded and cancelled orders: you did not keep that money, so it is not revenue — we still rendered the cart, so it is usage.

Limitations

These are real gaps, not things you have configured wrongly.

No figure on this screen is measured yet. Attributed revenue, orders that used the cart, revenue by module, both average order values and the funnel are all waiting on the two sources described above. The "as of" line currently reads that the figures have not been calculated yet.
Average order value needs more than events. Even once storefront events start arriving, the two average order value cards and the "% of orders" share need your store's total order count, which this screen does not yet have. They will keep saying "No data yet" until that work is done.
The funnel is a placeholder. It is a single "Cart opened to checkout" card waiting on data, not a step-by-step funnel you can break down.
The period buttons do not filter the figures yet. Choosing a period changes the labels on the screen and in the exported file, but the totals behind them are not yet restricted to that period. Once real numbers arrive, treat the period as a label until this is fixed.
There is no date picker on the screen. Only the four preset buttons are shown. Custom ranges are understood by the page but there is no field to type them into yet.
Forwarding to Google Analytics 4 and Meta is not available. It does not currently work — each destination rejects or discards what is sent because required fields are missing — so it has been taken off the pricing page and should not be counted on. Do not plan reporting around it.
Nothing here is a Shopify report. These figures describe the cart drawer's involvement in orders. Your Shopify admin remains the source of truth for sales.

If something looks wrong

A card that says it is waiting is working as intended. A card showing a number you cannot account for is worth reporting — and the published method above is the fastest way to check whether a figure means what you assumed.

Getting the cart in front of shoppers is what starts any of this measuring. See Publishing for taking a cart live, and Support if a number does not match what you expected.