How to use Shopify product exporter

The pass

Paste the store URL. Open Shopify product exporter and do it in this order.

Paste the store whose public assortment you want in a sheet.
Open the CSV and sort by price. Mark compare-at versus price, and count variants under a single title.
Ignore the absence of cost. That column is not supposed to be there.
Take any price you might match into the margin calculator before you change your own product.

The mistake that wastes the pass

Building a cost model from the CSV. The cost was never in it. You will be modeling a blank cell.

How you know it worked

Two live product pages match the prices in the sheet. You did not add a cost column from memory.

Save the file, then name the snapshot

When the download completes, you save it immediately into a folder you control, with the store domain and the date in the name. You do this because the file is not stored for you. Then you open the CSV and read the header row so you know which columns this export actually gave you. You confirm that you see titles, prices, compare-at, variants, SKUs, and images in whatever form the file uses. A header you have not read is a header you will misquote.

You add a note beside the file, a few lines in the same folder. The note repeats the store URL, the download time, and the fact that cost, drafts, and admin inventory are absent. Future you, and anyone you hand the folder to, needs those absences in writing. The CSV will not explain itself. A three-line note prevents a month of wrong assumptions.

Say the download opens in a spreadsheet tool that wants to reformat SKUs as numbers. You stop that reformatting if you can, because a SKU is a label. Leading zeros and mixed characters matter. If the tool already changed them, you note the damage and you re-export before you rely on the SKU column. Labels that have been "helpfully" converted are a common way a clean catalog becomes nonsense.

Spot-check two live prices and write the verdict

You choose two products you are willing to talk about later. They should be rows you might cite, not two random rows you will never mention. For each one you open the live product page on the store and you find the matching variant. You compare the sheet price with the live price. You compare compare-at the same way when both the sheet and the page show it. You write matched or differed for each product, in the note next to the file.

Two checks are the whole assignment. You do not expand it into a claim that every row matched because the first two did. You do not hide a mismatch on the second product because the first was clean. If one differs, the sheet is unverified as a price authority. You may still use titles and images with care. You quote a price only when that specific price was checked, or you label it unchecked and dated.

Say the first product matches on title and price, and the second product matches on title while the live price differs. You write both sentences. You do not invent a reason that flatters the store or flatters your download. You can re-export and check again, or you can cite the live page for the second product. The pass is the written verdict, and the verdict is allowed to be mixed.

Mark the columns you will cite and the ones you will ignore

In the note, list the columns you intend to cite this week. Typical cites are title, price, variant name, and image. Then list the columns you will ignore in conversation, starting with anything you wish were there and is not. You ignore cost because it is absent. You ignore stock because admin inventory is absent. You ignore the temptation to treat a SKU as proof the variant is available to buy in quantity.

Compare-at gets a rule. You cite it only as a public pair with the price, and only for rows where you still need it. You attach the phrase that it is a presentation on the storefront. Anyone who wants to say more has to bring evidence the CSV does not contain. Your rule keeps a sortable column from becoming a speech.

Images get a rule too. You may look at them to understand the product. You check rights before you reuse them anywhere public. A URL in a spreadsheet is not a license. For internal recognition, looking is often enough. For your own Shopify theme, you plan on assets you have the right to upload.

Hand off the snapshot with the checks on top

The handoff packet is the saved CSV, the three-line note, and the two price verdicts. You put the verdicts at the top so a hurried reader sees them. You name the owner of the next step. Honest next steps include mapping an assortment against your own catalog, asking a supplier about a similar item, or archiving the file as reference. Dishonest next steps include pasting their prices into your admin or building a purchase order from their SKUs.

Say your actual question is which products shoppers are ordering, by country and category, rather than what this storefront publishes. That question belongs to the daily product list, which is a different record from a catalog snapshot. You may run that list as its own pass. You do not stuff ordering hopes into the CSV. The sheet does not know which of its rows shoppers are ordering. It knows what the store publishes.

You finish the pass by reading the note aloud once. Store URL, date, two verdicts, absent cost, absent drafts, absent admin inventory, next owner. If any sentence sounds like a profit claim, you delete it. The public catalog is now a file you can find again, with its limits stapled to it. That is the whole discipline.

How you pick the two products worth checking

The two live checks are only as useful as the rows you choose. You pick products a real conversation might cite: a hero you keep mentioning, a variant with a compare-at value, or a title close to something in your own catalog. You skip a random row buried in the sheet if nobody will ever say its name. A perfect match on a product you will not discuss does not protect you when someone quotes a different price from memory of the file.

You match variant to variant. Say the sheet lists a lamp in two finishes and the live page defaults to one finish. You check the finish you intend to mention, and you write that finish in the verdict. A match on brass does not clear the black row. SKUs help you keep the finishes straight when titles repeat. You still describe the SKU as a label you read, and you leave quantity unwritten.

You capture the verdict in a sentence a stranger can trust. Name the product, say whether the sheet price and the live price matched or differed, and say the same for compare-at if you looked. Add the time of the look. Two sentences of that kind, placed above the file name, are the pass. A paragraph of impression is not a substitute. If you cannot write the sentence, you have not finished the check, and the folder waits.

What you deliberately leave out of the folder note

You leave out guesses about why a price differed. A difference is enough to retire that cell from quotation. Causes can include an edit after the snapshot, a variant you mismatched, or a page that shows a different currency presentation. You may investigate if the row matters. You may also simply cite the live page and move on. What you avoid is a confident cause you did not verify, because a wrong cause travels further than a humble mismatch.

You leave out any column you wish the exporter had added. Writing "stock unknown" is useful. Writing a guessed unit count is poison. Writing "cost unknown" is useful. Writing an assumed cost from a category rule of thumb is poison. The note can list unknowns. It cannot decorate them. Drafts get the same treatment: you can say unpublished products are outside the file, and you cannot name products you believe they are hiding.

You leave out a rank of the store. A spreadsheet invites sorting, and sorting invites a podium. Public prices sorted from low to high are still public prices. They are not a measure of who is winning the category. When the note is free of podiums, causes you invented, and filled-in blanks, the folder is ready for whoever owns the next step.