How to Build a Shopify Total Sales Trend Export That Can Explain a Dip
Build a Shopify total sales over time export that explains a dip: pick the right time grain, add the fields that diagnose it, validate it and schedule it.

When you export Shopify's total sales over time, the time grain and the fields you include decide whether the export can explain a dip or just report one. A monthly export tells you something changed in July. An hourly export tells you a flash sale drove the first spike and a payment outage caused the second drop. This guide walks through a reusable export spec in Data Export: how to pick the grain, which fields turn a trend into a diagnosis, and how to schedule it so your team gets the same clean report every week.
The focus here is the export workflow, not a tour of Shopify's native report. If you want a plain description of the Shopify UI report, that is covered in Shopify's own help documentation. If you want a repeatable file that lands in your inbox every Monday with the columns you need to act on, keep reading.
What does "total sales over time" actually measure?
Total sales over time is the same money figure Shopify shows everywhere else in Analytics, split across a time axis instead of summed into a single number. The formula is:
Total sales = Gross sales − Discounts − Returns + Shipping + Taxes + Duties
The time axis is what makes this view useful. A single period total of $84,320 tells you nothing about whether the store is healthy. The same $84,320 spread across 30 days, with one day at $7,100 and one day at $380, tells you something happened on the quiet day and gives you a point to investigate.
Two things trip people up here. First, this view uses order date, not payout date. A sale placed on October 29 shows up on October 29 in this trend, even if the money hit your bank on November 4. Second, returns are subtracted from the day the return was processed, not the day of the original order. That means a refund issued in November can pull November's total down even though the original revenue sat in October's row.
Which time grain should you pick for a trend export?
The grain depends on the decision the report has to support. Pick one deliberately, because the right grain for a quarterly board slide is the wrong grain for diagnosing a flash sale.
Hourly: use it for one to seven days around a known event. Launch days, flash sales, server incidents, ad bursts. Anything shorter than a day is noise for a weekly report, but it is essential when you need to see a two hour drop inside a Monday.
Daily: use it for a standing weekly or monthly report, and for most ad hoc investigations. One row per calendar day is the grain that answers "which day" without drowning the sheet.
Weekly: use it for quarter long reviews where day to day volatility hides the trend. A weekly view is also better for merchants whose traffic is heavy on weekends, because week over week removes the Monday vs Sunday effect.
Monthly: use it for year over year planning, finance reviews, and anything above a quarter. Monthly rows smooth out promotion timing and compare cleanly across years.
A practical rule: match the grain to the shortest decision window. If you will act on the report within a week, daily rows carry enough signal. If the fastest possible action is a monthly planning meeting, daily rows are usually clutter and weekly or monthly is better.
Which fields turn a trend into a diagnosis?
A trend export with only Date and Total Sales will tell you when something changed. It will not tell you why. To diagnose a dip without reopening the Shopify UI for every data point, include the components that go into the total sales formula, plus a few shape metrics.
Core trend columns
Date (or Date and Hour): the grain you chose above.
Order Count: the number of orders placed in that row. A dip with flat order count points at pricing or discounting, not traffic.
Gross Sales: product price multiplied by ordered quantity, before anything else. This is your top line.
Discounts: the amount taken off through discount codes and automatic discounts. A spike here explains a soft total even with healthy orders.
Returns: the value removed from sales when refunds were processed on that day. This is the field that moves without a corresponding order movement.
Shipping, Taxes, Duties: add the three that apply to your store. They complete the math and let you reconcile the row against Shopify.
Total Sales: the output of the formula. Keep it as a column so a reader can sort the sheet on it.
Shape columns that explain the total
Average Order Value: Total Sales divided by Order Count, computed in the report. A falling AOV with steady order count is a different story from falling order count with steady AOV.
Discount Rate: Discounts divided by Gross Sales. Cleaner for trend watching than raw discount dollars, because it is comparable across busy and quiet days.
Return Rate: Returns divided by Gross Sales for the same row. Flags days when refund processing, not new sales, moved the total.
Add the shape columns as calculated fields in Data Export rather than leaving them to a spreadsheet formula downstream. That way the scheduled file arrives ready to read, and the person opening it at 8 a.m. does not have to re edit a cell.
How do you build the export in Data Export?
Here is the setup for a daily trend. Swap the summarization step for hour, week, or month to produce the other grains.
Open Data Export and go to Reports, then click Create a Custom Report. Pick Sales as the category.
Select Order Date as the primary date field. This matches Shopify's native Total Sales Over Time report. Payout date and fulfillment date are separate decisions covered later.
Add the columns from the field list above: Order Count, Gross Sales, Discounts, Returns, Shipping, Taxes, Total Sales. If your store collects duties, add Duties too.
In Report Builder, use Summarize By to group rows by Day. The pre made Sales options also include Hour, Week, Month, Quarter, and Year. For a daily trend, pick Day.
Add two calculated columns. Average Order Value: Total Sales divided by Order Count. Discount Rate: Discounts divided by Gross Sales. Format them as currency and percentage.
Set the date range. For a standing weekly delivery, use a rolling window such as Last 7 Days or Last 30 Days. For a quarterly review, use Last 90 Days.
Apply filters only if you need a slice. To build a baseline trend, leave the channel and location filters open so the export matches Shopify's totals. To compare channels, clone the report and filter each copy by Sales Channel.
Freeze the Date column and sort descending so the most recent row sits at the top of the file. Save the report with a clear name such as "Total Sales Daily Trend, Rolling 30d."
How do you validate the export before scheduling it?
Before you schedule a report that lands in a shared inbox, confirm it matches Shopify for a known window. Three checks catch almost every setup mistake.
Match a closed day. Pick a day that is at least three days in the past, where refunds and edits have settled. Compare Total Sales in your export to Total Sales in Shopify's Analytics for the same single day. The two numbers should match to the cent. If they do not, the most common causes are a time zone mismatch, a channel filter applied by accident, or the export running on Created At rather than Order Date.
Check the math inside one row. For any row, confirm Gross Sales − Discounts − Returns + Shipping + Taxes + Duties equals Total Sales. If the row does not reconcile, a column is missing or is being summarized on a different basis than the rest.
Reconcile the window total. Sum Total Sales across the whole export and compare to Shopify's total for the same window. Small differences usually come from the time zone the export uses versus the store's reporting time zone. Fix the time zone once in the report settings.
Document what the report matches and what it does not. A short note at the top of the Google Sheet, such as "Order date, store time zone, all channels, all locations," saves an hour the next time somebody asks why the number is different from a different dashboard.
How do you schedule the export so it arrives every week?
Once the report reconciles, put it on a schedule so no one has to re run it. In Data Export, open the saved report and click Schedule.
Cadence: weekly on Monday morning is the common default for a daily grain export. For monthly rollups, the first business day of the month works well.
Format: Excel or Google Sheets for analysts who will filter and chart it. CSV for anyone piping it into another system. PDF for stakeholders who only need to read the numbers.
Delivery: email for individual readers, Google Drive or FTP for files that feed another tool. Google Sheets delivery writes the new data into the same sheet every week, which keeps your charts pointing at one URL instead of chasing new files.
Recipients: add everyone who needs to see the file, not just the person who requested it. A sales trend used by only one person often gets re explained to three others over the week.
For sensitive reports, add password protection on the file before delivery. This is useful when the schedule emails to a shared distribution list and you want only intended viewers to open it.
When should you build a second trend export instead of filtering the first?
Filters inside a single export are tempting because one saved report feels tidier than three. In practice, the baseline trend and the sliced trends have different jobs.
Keep the baseline export unfiltered. It is the reconciliation report, the one that has to match Shopify's own total. Any filter applied to it makes future reconciliation harder, because a reader has to remember the filter before comparing.
Build separate, clearly named exports for slices: Total Sales Daily by Channel, Total Sales Daily by Location, Total Sales Daily by Customer Tag. Each one answers a specific question, and each one can be scheduled to a different audience. A store manager wants the location slice. A paid media lead wants the channel slice. Collapsing all three into one file with a dropdown is clever, but it slows everyone down.
What about date fields other than order date?
The trend export above uses order date because that is what matches Shopify's native total sales over time. There are two cases where a different date is correct, and they belong in different reports, not in this one.
Payout date: use this when you are reconciling to the bank. The money moved on the payout date, not the order date. A finance focused trend export should use payout date and should carry Shopify fees as a column.
Fulfillment date: use this when the question is about operations. A fulfillment date trend tells you when work left the warehouse, not when a customer placed the order. It will not match Shopify's Total Sales Over Time report, and that is fine, because it is answering a different question.
Keep these as separate saved reports in Data Export. Naming them distinctly, such as "Finance Trend, Payout Date" and "Ops Trend, Fulfillment Date," prevents a reader from pulling the wrong one into a meeting.
The short version
A total sales trend report is worth building once and running on a schedule. The decisions that matter are the ones you make before you hit Schedule. Pick the grain against the decision window. Include Order Count, Gross Sales, Discounts, Returns, Shipping, Taxes, Total Sales, plus Average Order Value and Discount Rate as calculated columns. Validate against a closed day before anyone else sees the file. Keep the baseline unfiltered and build separate reports for channel, location, or customer slices. With that setup, the Monday morning file is the start of a decision, not the start of a reporting task.
Frequently Asked Questions
Does the total sales figure in my export match what Shopify shows in its Total Sales Over Time report?
It should, to the cent, as long as the export uses Order Date, the store's reporting time zone, and no filters on channel or location. If the totals are off, check those three settings first. A time zone mismatch is the most common cause of small differences, and a stray channel filter is the most common cause of large ones.
Why should I export on order date rather than payout date for a sales trend?
Order date answers the sales question: when did customers buy. Payout date answers the finance question: when did money reach your account. The two will never match, because Shopify payouts arrive on a lag and group multiple days of orders into one deposit. Use order date for trend analysis and build a separate payout date report for bank reconciliation.
Can one scheduled export cover hourly and monthly grain at the same time?
No, and you would not want it to. One scheduled file should answer one question at one grain. Build an hourly export for the window you need it, usually a specific launch week, and keep the daily export running on its weekly cadence. Mixing grains in one file makes the file harder to chart and harder to explain.
What should I do if two days in my trend look identical but one is actually off?
Look at the component columns in the same row. Two days can share a Total Sales number while hiding very different underlying activity, for example one day with high gross sales and heavy discounting, the next with lower gross sales and very few refunds. That is why the export carries Gross Sales, Discounts, and Returns as separate columns, not just the net total.
Can I split the trend by sales channel in the same export?
You can, but a single file with a channel breakdown becomes hard to read at daily grain. The cleaner pattern is to keep one unfiltered baseline export that matches Shopify's totals, and add separate saved reports for each slice: one for Online Store, one for POS, one for other channels. Each slice can be scheduled to the audience that needs it.
Try Data Export free for 15 days
Custom reports, 100+ prebuilt reports and scheduled delivery on every plan, from $12 per month.
