Compare this year to last year on the Trends tab

Ecommerce › Trends is your store's own history: every month of sales laid out year against year, and every product ranked by how many you actually sold. It answers the two questions a spreadsheet is usually opened for, "what does November normally look like" and "which of these is not moving any more", without opening one.

The month grid

Years run down the left, January to December across the top, and the shading gets stronger with the number, so a busy season shows up as a band of colour before you have read a single figure.

The four buttons above it change what every cell means:

• Revenue is what the products sold for, after any discount, before tax and before shipping.
• Units is how many items went out the door.
• Orders is how many orders came in.
• Avg order is revenue divided by orders. For a year it is the whole year's revenue over the whole year's orders, not an average of the twelve monthly averages, which would let one quiet January count as much as a busy December.

A blank cell is not a zero. Blank means the month is outside the history TALISK_HQ holds, usually because it has not been imported that far back. A month that was imported and genuinely sold nothing shows a real 0, so a quiet December in the middle of your history reads as a quiet December and not as missing data. The line above the grid always says which range you are actually looking at.

The product list

It opens sorted by units, because that is the honest ranking of what your customers keep buying. Last sold is the column most people end up using: scroll to the bottom and you are looking at stock that has not moved, with the date it last did.

Click any column heading to sort by it, and click the same one again to reverse it. A name sorts A to Z first; units, revenue, average price and last sold start with the biggest or the newest. A small arrow marks the column the table is currently ordered by.

The box above the table searches your products by name or SKU. It searches everything your store has ever sold, not just the rows on screen, so a product that sits too far down the list to be shown is still found by typing its name. While you are searching, the box says how many of your products matched.

The table shows the first 100 rows. If there are more than that, a line under it says so and tells you how many there are, and searching is how you reach the rest.

How many orders were over a certain size

Under the grid, Order size answers "how many orders were worth at least this much" for every month of every year. Type any figure into the box and the grid recounts, so you can move the line to whatever counts as a big order for you. Beside the box it tells you how many of your orders cleared it and what they came to.

The figure it starts on comes from your own orders. It opens on roughly your middle order size, rounded to a round number, and says so beside the count. That is so the first thing you see is a real answer rather than a column of zeros: a fixed starting figure only suits a shop whose orders happen to be that size. Type over it whenever you want a different line.

These figures are counted differently from everything else on this tab, and the note under the grid says so. An order's size here is what the customer actually paid, tax and shipping included, because that is what "an order over $1,000" means. The revenue figures further up are the products alone, before tax and shipping. So the two will not add up to each other, and neither is wrong.

Three more things it does, all deliberate. Gift cards are left out, the same as everywhere else on this tab. A refund does not un-count an order, so an order of $1,200 that was later refunded was still an order of $1,200. And if an order was edited before payment was taken, it is counted at what it ended up being, not at what was first placed.

Underneath, the mix, year by year puts your orders into price bands and shows how many fell in each, for every year. Beside the count is that band's share of that year's orders, which is the number to read if you want to know whether your mix is shifting: a year where you simply sold more will show bigger counts and the same shares, while a year where you sold bigger shows the shares move.

The band edges are yours to set, in the box under the table. They are not worked out from your sales on purpose. Bands that moved with the figures could not be compared from one year to the next, and comparing years is what the table is for.

This section reads your orders directly rather than the monthly summary the rest of the tab uses. Both keep up with your store as orders come in, so they should agree on how many orders a month held.

How many of one product did I sell each month

Click a product and the same month grid opens in the row underneath it, counting units of that one product, so you can see at a glance whether it is seasonal, growing, or simply finished. Click the row again to close it. The arrow at the left of each product name is what marks a row as openable.

More than one store

If your business has more than one storefront connected, the store selector at the right of the Ecommerce tab bar narrows Trends to one store: the month grid, the product list, the order-size count and the line saying how much history was imported are all that store's. Pick All stores to add them back together.

Why the numbers here do not match your books

They are counted from a different thing, on purpose.

Your Ecommerce overview and your profit and loss are built from money that has actually landed and been posted to your ledger. Trends is built from the orders themselves. An order placed in December and paid out in January is December here and January there. Sales tax, shipping and refunds are handled differently again.

So the two will not tie, and Trends is kept on its own tab with its own range for exactly that reason. Use Trends to see shape and seasonality; use the overview and your reports for what the business actually earned.

Why there is no margin here

TALISK_HQ records what a product cost you at the moment its order syncs. An order imported today from three years ago would be stamped with today's cost, so a margin worked out across your history would look precise and be wrong. Margin lives on Ecommerce › Products, over a period where the cost is the one that applied.

Bringing your history in

At the top of the tab there is one control. Pick how far back you want to go, 1, 2, 3 or 5 years, and press Import history. Before you choose, the screen tells you the exact range it will actually fetch for that choice, so you are never picking a number and finding out afterwards.

It runs in the background and you can leave the page. The newest months arrive first, on purpose: this year fills in within minutes while the older years are still coming, so there is something useful to look at straight away. The progress line says how many months are done and how many orders have arrived.

If your store has been connected to TALISK_HQ for a long time, you may not see that form at all. A connected store syncs its orders as they come in, so your grid can already reach back further than any import does. When that is the case the tab says so in a line instead, naming the month your history starts, and the import form is folded away behind Import anyway. It is folded rather than removed because an import can still be useful: it re-reads the window, which fills in anything that went missing. What it cannot do is reach back past what you already have.

It does not change your books, your reports, or your tax figures. Only past orders and products come in. Past payouts are deliberately left alone, because those are the entries that post to your ledger, and pulling in years of them would push sales into periods your accountant has already closed.

Adding another year later

Once an import has finished, the form folds itself away. In its place is a single line, Go further back; open it and the same range row is there again. It carries on from the oldest month you already have, so nothing you already imported is fetched twice, and you can only go further back, never over the same ground again. If an import stops before it finishes, the screen says why and offers to pick it up where it left off.

The ranges you already have are still on the row, greyed out, so you can see what you took as well as what is left. The selection starts on the next one that is actually available: after a one year import, 2 years is highlighted and 1 year cannot be pressed.

When there is nothing further to ask for, the control is gone and the tab says so in one line instead. That line is about how far back an import can REACH, not about which months you ended up with. The range you are actually looking at is always on the line underneath, and on a store limited to the last 60 days those two are different.

Free accounts can import one year. Any paid plan goes back five.

How far back it goes

Shopify only lets an app read the last 60 days of orders until it grants extended access, and TALISK_HQ has that request in. Until it is granted, the control tells you so before you choose, and an import can only reach back that far.

That limit is about what an import can FETCH, not about what this screen holds. If your store has been connected to TALISK_HQ for a while, its orders have been syncing all along, and every one of those months is already on the grid. A store connected two years ago will see two years here whether or not the extended access has been granted. What the limit costs you is the years from BEFORE you connected. When the access is granted, Go further back is how you go back and get them.

The same thing works the other way round for a store that has been connected for years: its synced history can start earlier than the longest import reaches, and then there is genuinely nothing older to fetch. That is what the line about your history starting in a given month is telling you.