The short version: The trap is that the image on your old storefront is not your original. Platforms resize and recompress on upload, so downloading from your own product page gives you a degraded copy that will look worse on the next platform, and worse again on the one after that. Find the originals on your phone or camera first. Descriptions move as plain text and lose their formatting, which is annoying and also the best opportunity you will ever get: migration is the only time anyone re-reads all forty product descriptions, and most of them are worse than their author remembers.
Because what you uploaded is not what is being served.
Almost every platform processes images on upload: resizing to fit its templates, compressing to load quickly, sometimes converting the format. That is sensible for page speed and it means the file on your product page is a derivative, not your file.
Download that derivative, upload it to a new platform, and it gets processed again. Two rounds of compression on the same image is visible, particularly on food, where the detail in a crumb or a glaze is exactly what compression discards.
So the order of preference for every image:
Most vendors have option one and do not think to look, because the storefront is where the photograph lives in their mind.
Work backwards from when you took them.
If you genuinely cannot find an original, that is a signal in itself: that product needs a new photograph, and taking one is a better use of an hour than salvaging a compressed copy. Our guides to product photos with only a phone and food photography for market vendors cover doing it well without equipment.
Going forward, keep a single folder of originals, named to match your products. It costs nothing, it survives every platform change, and it is the asset you will still be using in five years.
They move as text and arrive stripped.
Expect to lose: bold and italic formatting, bullet lists, headings, links, and any spacing you were relying on. What survives is the words. On some platforms an export gives you HTML instead, which preserves formatting and is harder to paste into a different system cleanly.
The practical approach:
Step two is where the value is. Descriptions accumulate: one written carefully when you launched, the next twenty typed at 11pm, several that still say "coming soon" for a product you have sold two hundred of. Nobody notices because nobody reads their own store.
Six things, and most food listings have two of them.
Our guide to writing product descriptions that sell food online covers the structure, and our piece on farm stand product descriptions covers the shorter version.
Point six is the one most vendors under-invest in, and it is the only one doing any selling. Point one has consequences beyond conversion. An allergen question answered on the page is an allergen question you are not answering by text with floury hands, and the FDA's food allergies guidance covers the major allergens that belong in every listing.
You do, if you took them, and it is worth knowing what that means before you move them anywhere.
A photograph you take is your copyright from the moment you take it. Uploading it to a platform generally grants that platform a licence to display it, and does not transfer ownership. So moving your own images to a new platform is entirely yours to do.
Where it gets complicated:
The Copyright Office's registration guidance covers what protection you have by default and what registering adds. For most food vendors nothing needs registering; the useful point is simply that your own photographs are yours, and any platform suggesting otherwise deserves a careful read of its terms.
They break, and this catches people who have shared images anywhere.
If you have ever linked to a product image directly, embedded one in an email, or used a platform-hosted image on another site, those links point at the old platform's servers. When you cancel, they stop working, and the email you sent last month now has a broken image in it.
What to do:
This is a small job that becomes impossible at exactly the wrong moment.
Four things, all small, all irritating to discover after the fact.
Product options and variants. A cake with three sizes and four flavours is twelve combinations, and platforms structure that differently. Some use variants with their own prices and stock; some use add-ons that modify one product. They do not map cleanly onto each other, so plan to rebuild rather than transfer, and check the pricing comes out right on the most complicated item you sell.
Sort order. Your best seller was at the top because you put it there. A fresh upload usually sorts alphabetically or by creation date, and nobody notices until sales of the thing that used to be first quietly drop.
Product URLs. Each listing had an address, and those addresses appear in old posts and in search results. They change on the new platform unless you can set them, and links to them break.
Image orientation. Photographs taken in portrait sometimes upload rotated, because orientation lives in metadata that gets stripped during processing. Check every image after upload rather than assuming.
None of these is a reason to stay where you are. They are the reason to migrate in one block and then look at every product page on a phone before you announce anything. That check takes fifteen minutes for forty products and catches all four. If you are partway through evaluating somewhere, upload your most complicated product to a trial first rather than your simplest, since the simple one will work everywhere and tells you nothing.
Less than people fear, if you sequence it properly.
For a catalog of twenty to forty products:
So roughly a weekend, and the descriptions are the part that pays you back rather than the part that costs you.
The mistake is doing it in fragments across three weeks while also running the business. Block one weekend, get the whole catalog across, test it, and then switch. Half-migrated is the worst state to be in, because you are maintaining two systems while neither is finished.
If a weekend is genuinely not available, the workable alternative is to migrate by product line rather than by task. Get your five best sellers completely finished, photographs, descriptions, options, prices, stock, and pickup rules, then stop. Those five are sellable. Next session, do the next five.
That works because a finished product is useful and a half-finished catalog is not. What fails is doing all the photographs one week, all the descriptions the next, and all the prices the week after, because at no point during that do you have anything you could actually sell.
Almost certainly, and it is the least discussed benefit of moving.
Most catalogs contain items that sell once a quarter, exist because someone asked for them in 2024, or duplicate something else with a slightly different name. Each one costs you a photograph, a description, a stock decision every week, and a line on your pick list.
Migration forces you to touch every product, which makes it the natural moment to ask of each: would I bother building this from scratch today? If the answer is no, do not move it.
A vendor who moves 40 products and keeps 28 has a faster migration, a cleaner store, a simpler bake, and almost never notices the missing twelve in the revenue. Our guide to batch economics and cost per unit covers how to tell which ones are genuinely carrying their weight.
Our guide to farm stand product photography covers batching new shots efficiently if you decide several products need fresh images, which is a common and correct conclusion.
A folder you own, containing:
That folder is your catalog. The platform is just where it currently lives. Vendors who keep it find their next migration takes an evening rather than a weekend, and vendors who do not end up rebuilding everything from scratch each time they move.
If you are choosing where to move to, Homegrown is $10 a month billed annually with 0% commission and 2.9% plus $0.30 processing published up front, and it handles pickup at each place you sell with its own schedule and cutoff, local delivery with a radius and a route, and sales tax calculated, filed, and remitted in all 50 states. The honest bounds: it does not ship nationally, it is not a website builder, and there is no bulk product import, so a large catalog goes in by hand. For twenty products that is an evening; for two hundred it is a real consideration and Big Cartel's bulk import or Shopify's CSV upload would save you time. You can load five real products with their photographs during a trial and see how the images render before committing to the whole catalog.
Every platform below shows the same four commercial facts, because a table that lists one platform's transaction fee and not another's is not a comparison. "Not published" means exactly that: the company does not state it publicly.
| Platform | Getting your catalog out | Subscription (annual) | Free trial | Platform fee | Card processing |
|---|---|---|---|---|---|
| Shopify | CSV product export including image URLs | $29/mo Basic | 3-day trial, then $1/mo for 3 | 2% platform fee if not on Shopify Payments | from 2.9% + $0.30 processing |
| Square Online | Item export from the item library | Free tier; paid from $29/mo per location | 30-day trial on paid plans | $0 platform fee | 3.3% + $0.30 free tier, 2.9% + $0.30 paid |
| Big Cartel | Product export available | Platinum $12/mo ($144/yr) | 7-day free trial | $0 platform fee | Your own provider, so 2.9% + $0.30 typical |
| Etsy | Listing download, images pulled individually | No subscription | n/a | $0.20 listing + 6.5% commission | 3.0% + $0.25 processing |
| Homegrown | No bulk product import, you add items by hand | $10/mo billed annually | 7-day free trial | $0 platform fee (0% commission) | 2.9% + $0.30 processing |
| WooCommerce | Full CSV import and export built in | $0 plugin + hosting | n/a | $0 platform fee | Your chosen gateway, typically 2.9% + $0.30 |
You can, but they will be resized and compressed versions rather than your originals. Uploading those to a new platform compresses them again, which is visible on food. Find the originals on your phone first.
Usually your camera roll, searched by the date you added the product. Also check editing apps, cloud backups that keep full-resolution copies, and any email where you sent them to yourself.
Generally no. Bold, italics, bullet lists, and headings are usually stripped when descriptions move as plain text. You will be re-entering them, which is a good reason to rewrite rather than paste.
You do, if you took them. Uploading grants a platform a licence to display them, not ownership. Check your agreement if a photographer took them, and check licence terms on any supplier or stock images.
They break when you cancel, because they point at the old platform's servers. Check old email campaigns and any other site where you have embedded a product photo, and re-host from your own files first.
About a weekend for twenty to forty products. Roughly two hours gathering originals, three rewriting descriptions, four uploading, and one setting pickup and delivery rules.
Yes. It is the only time anyone reads all their product descriptions in one sitting, and most stores contain at least one listing still saying something that stopped being true a year ago.
The important thing is the one nobody expects: the image on your storefront is not your original. Platforms compress on upload, so re-uploading a downloaded copy compresses it twice, and food photography is where that shows. Find the originals before you start, and if you cannot, take new photographs instead of salvaging.
Descriptions arrive stripped of formatting, which forces you to re-enter them. Treat that as the opportunity it is. Read all of them in one sitting, which you have never done, and rewrite the ones that are worse than you remember.
Then keep a folder of your own: original photographs, descriptions as plain text, prices in a spreadsheet. That folder is your catalog, and the platform is only where it happens to live this year.
