Shopify winning products: a worked Shopify study

What you would do

You filter to a country you ship, a category you can support, and a price band you could advertise. You open one row and write down the price. You do not order samples yet. You put that price, your product cost, your shipping, and a conservative ad cost into the margin calculator.

If the net margin cannot survive a bad week, you close the row. The board refreshes daily, so rejecting a candidate does not need a meeting. If it survives, you look at live ads for a similar item with the ad spy, still without treating those ads as proof.

Before you change the store

These are examples. Run the tool on your own URL or your own costs. That result is the one to act on. Do not put an example in a deck as if it were your store.

Example: one candidate row and the name on it

Say the daily list is open on a morning you set aside for candidates. You set a country you actually ship to, a category that matches the kind of goods already in your Shopify catalog, and a USD price band you are willing to look at. In that slice, one row is pet grooming glove. You write the date and the three filters before you let the title interest you. The row is a candidate because it is on the list after those filters, which is the only status the board is offering.

Say you have seen similar goods before and your first impulse is to treat pet grooming glove as a sure add. You leave that impulse out of the note. Shoppers ordering a product in this feed are the reason a row exists. That fact is about the list. It is not a statement about your theme, your traffic, or your refund policy. You are going to open the product, describe it, and stop at a candidate file.

You keep your admin available only to check for a name collision with something you already sell. Say you search your own catalog and you do not find the same product. You write "not currently in the catalog" as your observation. The board did not tell you that. Your search did. The two sources stay labeled so the example can be reread without confusion.

What you see when the candidate link opens

Say the link for pet grooming glove opens a product page that shows the title, a set of images, a visible price, and a short description of what is included. You paraphrase the inclusion in your own words and you quote only a phrase you might need later, such as a material the page names. You write that a price was visible. You do not copy the moment into a claim about a discount working, because a single page view has no history attached.

Say the link you used may earn a commission. You write that flag beside the URL. You also write that the commission does not make the row a sponsored ranking, and it does not prove the product will sell for you. Those two reminders sit in the example so the fee cannot hide. You are allowed to use the link. You are recording the economics of the click in the same note as the merchandise.

Say the page offers variants, such as two sizes. You name the sizes in the note and you do not assume both are something you can source. Variants on a public page are a description of that offer. Your future SKUs, if you ever create them, will be decisions in your admin. For this example, pet grooming glove remains a candidate with a page description attached, plus a commission flag, plus blanks.

The blanks you leave under that candidate

You leave unit cost blank. Nothing on the row or the product page is your supplier invoice. You leave shipping zones blank, then you add a separate line from your own memory of your settings only if you verify it in admin. Say you verify that the country you filtered is a country you ship to. You may write that verification, sourced to your settings, under the candidate. The board still did not know your zones. You did the check.

You leave the proof line blank on purpose. You do not write that pet grooming glove will sell, should sell, or has sold for stores like yours. You do not invent a review count, an order count, or a lift. The daily list already said its piece by including the row among products shoppers are ordering. Your note repeats the filters and the title. It does not upgrade the row into a case for a launch.

Say a teammate looks over your shoulder and wants the candidate moved onto a launch list before Friday. You show them the blanks. The path to a launch runs through a supplier quote, a sample, a margin check that uses your cost, and a page you are willing to publish. None of those appeared in the example. The labeled row stays in the candidate file with the date of the list on it.

How the example gets filed

The file has five lines anyone can audit. The date of the list. The country, category, and USD price band. The title pet grooming glove. The link and the commission flag. The blanks for cost, zones, and proof, plus your own catalog search result. A sixth line names you as the owner and names the next review, which in this example is a supplier question and not an ad brief.

Say the supplier question comes back unknown. The example does not fill the unknown with optimism. Unknown remains the status, and the candidate remains unlaunched. Say the supplier question comes back with a quote. That quote enters a different note, where margin can be calculated from numbers you possess. This example file keeps the pointer to the quote and continues to call the product a candidate until you deliberately change its status.

You do not add a creative script, a media budget, or an audience to the example. Those additions would turn a board row into a different kind of document. The row was a candidate on a daily list. The file you keep is the receipt of how you looked at it. Anyone who wants more has to bring a new source and name it.

The candidate the next afternoon still has not become demand

Say you open the file the next afternoon because someone wants a decision. The date, the country, the category, and the USD price band are unchanged. pet grooming glove is still the title you wrote down. The commission flag is still beside the link. The catalog search still says the item was not in your Shopify admin at the time you looked. None of those lines matured overnight into a sales figure. You do not update them to sound more certain just because a meeting is on the calendar.

Say the meeting asks for a sense of how big the opportunity is. You answer with the shape of the note: one row, one slice, one page you described, blanks where cost and proof belong. You offer to request a quote if the room wants a next step that creates new evidence. You decline to supply a forecast, a review tally, or a comparison with "typical" stores. Those requests are understandable. They are also outside the example you filed.

Say a later daily list shows pet grooming glove again under the same filters. You may add a second dated line that the title appeared on another day. Repetition is a fact about the list across two dates. It is still not proof the product will sell for you, and it is still not a sponsored rank. You keep both dates. You keep the blanks. The example stays a candidate file with a longer history of sightings, which is more honest than a single excited morning and also more modest than a launch.

The file can sit for a week without going stale in its claims, because its claims were small. Filters do not rot. A commission flag does not rot. A blank does not become a number by aging. What can go stale is your memory of the product page, so if you cite a page detail after several days you reopen the link and you add the new date. The original sighting remains the original sighting. The example is built to be reread by someone who was not in the room, which is the standard you hold it to.

Shopify questions