Weebly to Shopify Migration Case Study: How SoSlanty Moved 1,685 Products, 3,412 Orders and Its Google Shopping History in 44 Days

By Lance Allison, Founder of Ruby Digital Agency | September 2026 | ~41 min read

If you have ever needed a gift for the person who has everything and finds most of it boring, you need to know about SoSlanty. It is a one-woman gift shop in Northeast Ohio that sells, by my count, the funniest socks on the internet, and it does not stop at socks. There are coasters that say what everyone at the table is thinking, kitchen towels and oven mitts with opinions, cocktail napkins, boxed matches, magnets, flasks, enamel pins, journals and a women-empowerment line that is genuinely lovely. The women’s crew socks alone run to 168 designs, from “Always Be Yourself Unless You Can Be a Unicorn” to “I’m Not Bossy, I’m the Boss,” and the men’s gifts include a Christmas sock that reads “Hark! To the Microbrewery At Once” and a pair labeled, with admirable honesty, “Worst Gift Ever.” A good number of the bestsellers cannot be quoted on a business blog, which is rather the point.

It is also, quietly, one of the best holiday shops I have worked on. Most of the catalog sits under fifteen dollars, which makes it a stocking-stuffer machine: socks for everyone on the list, a coaster set for the office exchange, a gift bag for the wine bottle with a better joke on it than the card, boxed LGBTQ holiday cards (“Deck the Halls With Fabulous,” “May Your Yuletide Be Gay”), and a gift card for the impossible ones. Orders ship the same or next business day, shipping is free over $95, and the return policy runs a full 365 days, which tells you how confident the owner is that the recipient will laugh. If you finish this article and buy nothing, I have failed at the easy half of the job. Go and look at the new arrivals first.

The hard half of the job was the migration. In July the owner sent us a one-page brief with two questions near the bottom that I have been thinking about ever since. The first was “How many Weebly to Shopify migrations have you executed?” The second was “Any particular issues or quirks to look out for due to Weebly being so old?” They are exactly the right questions, and the honest answers are “not many, because almost nobody leaves Weebly in an organized way” and “yes, more than you would expect.” This is the story of that project, SoSlanty’s move from Weebly to Shopify, told with the numbers, the mistakes and the fixes, because a case study you cannot check is just an advertisement.

The short version: SoSlanty had spent nine years on Weebly, building up 1,658 products, 3,412 orders and a Google Shopping feed that was quietly doing a lot of the selling. Weebly, owned by Square since 2018, has stopped developing and is now being wound down in 67 countries. We moved the store to Shopify in 44 days from kickoff to go-live. The entire migration, products, customers, orders, redirects, reviews and the Google cutover, was run on our own AI-driven migration technology: purpose-built scripts executed against Shopify’s Admin API, with a person checking every category of data against the source before launch. Then we spent a further two weeks moving the Google side without losing the history, which turned out to be the hardest part of the whole job. This article covers all of it, and closes with the lessons that apply to any Weebly or Square Online store thinking about the same move.

For anyone arriving from a search: a Weebly to Shopify migration is the process of moving an online store’s products, customers, order history, URLs and marketing integrations from Weebly (now run on Square Online’s ecommerce back end) to Shopify, then rebuilding the storefront on a Shopify theme, because design never transfers between the two platforms. This ecommerce migration case study documents one such move end to end, with every figure taken from the project files rather than from memory.

In This Article

1. The Client: Nine Years on Weebly and a Catalog of 1,658 Products
2. Why Leave Weebly in 2026 (Even If Your Country Is Not on Square’s List)
3. Where Weebly’s Data Actually Lives: Square Online, a Thin Export and a 1,000-URL Sitemap
4. Running the Migration on Our Own AI-Driven Technology: Products, Customers, Orders and the Checks Between
5. The Redirect Map: 1,758 Redirects for URLs Nobody Could Guess
6. Rebuilding the Storefront in Horizon: Menus, Sayings and 110 Rescued Reviews
7. Go-Live Day, Step by Step
8. Keeping the Store’s Google Shopping History
9. The Numbers: What Moved, What Was Verified and What We Did Not Measure
10. What This Ecommerce Migration Case Study Teaches About Any Weebly to Shopify Move
11. A Quick Overview of Ruby Digital Agency
12. A Ten-Step Plan for Your Own Weebly to Shopify Migration
13. What Happens If You Do Nothing
14. Frequently Asked Questions
15. Final Thoughts

The Client: Nine Years on Weebly and a Catalog of 1,658 Products

SoSlanty sells funny, affectionate, slightly rude gifts: socks with sayings on them, coasters, kitchen towels, oven mitts, cocktail napkins, boxed matches, magnets, mugs, flasks, enamel pins, journals, greeting cards, wall signs and a line of women-empowerment and LGBTQ gifts. The anchor brand is Blue Q, whose socks are the store’s main draw, alongside Twisted Wares, Tipsy Coasters, MatchDaddy, Magnetic Poetry, Lucky Jack and around twenty small independent artists. It is a one-person business run by its owner from the Cleveland area, with a 365-day return policy, same-or-next-business-day shipping and free shipping over $95. The tagline on the old site was “awesome stuff with creative images and bold thoughts,” and the footer said “Live Happy, Be True to Yourself.” That voice mattered to the rebuild more than any technical decision.

The store had been on Weebly since 2017. The orders export runs from August 17, 2017 to September 2, 2026 and contains 3,412 orders. The home page carried 110 customer reviews with a 4.93 average, collected through a third-party widget. Google Search Console had been verified by the owner in March 2023, GA4 had been running since February 2024, and a Google Shopping campaign had been feeding Merchant Center from a hand-maintained Google Sheet with product IDs in the form BB1 to BB1707. In other words this was not a hobby site; it was a small business with nine years of accumulated data and a search footprint worth protecting.

The brief said “about 1,900 items.” The export said 1,658 products, 1,666 variants and 8,973 units on the shelves. That gap between what an owner remembers and what the platform holds is normal and is the first thing a migration should settle, because scope, price and timeline all hang off it. Here is the catalog as it stood at kickoff and at go-live.

Metric	At kickoff (July 2026)	At go-live (September 2, 2026)	Notes
Products	1,658	1,685 active	27 products added during the build; the old gift-card placeholder archived
Variants	1,666	1,693	Mostly sock sizes and colors
Units on hand	8,973	Matched the owner’s export to the unit	8 shelf-count corrections applied July 30
Customers	3,187	3,213	26 new customers from the final delta sync
Orders	3,386	3,412	August 2017 to September 2026; 26 placed during the build
Collections	44	45+	Two categories split on the owner’s request
Images	One per product in the export	4,001 on Shopify’s CDN	Additional images harvested from the live product pages
Store reviews	110 on the home page	108 unique imported to Judge.me	Plus 112 product reviews typed up by the owner
Vendors	None recorded	32	Backfilled by hand; 8 products still awaiting a real brand

One more number sets the scale. Median price across the 1,693 variants is $13.00; 568 variants sell for under $10 and only 16 for over $30. This is a high-volume, low-ticket catalog where nobody is going to notice a missing product until a customer emails about a dead link. That informs the level of verification we applied later. If you want the general playbook before the specifics, our Weebly and Square Online to Shopify migration guide covers the mechanics; this article is what happened when we ran them.

Why Leave Weebly in 2026 (Even If Your Country Is Not on Square’s List)

Square bought Weebly in 2018 and folded its ecommerce into what became Square Online. Since then Weebly has been maintained rather than developed. New Square customers are steered to Square Online; Square announced that the Weebly mobile app would leave the app stores on December 1, 2025; and earlier this year Square published a support notice titled “Upcoming Changes to Weebly’s Global Operations.” The operative sentence reads: “We have made the difficult decision to wind down Weebly services in 67 countries by 2026-09-20 due to changes in regulation and to simplify our global operations.” The dates that follow are worth having in front of you.

Date	What Square’s notice says happens	Who it applies to
June 29, 2026	“You will no longer be able to publish any new pages”	Weebly customers in the 67 listed countries
September 20, 2026	Services “wind down” in those countries	Same
September 27, 2026	“Your current Weebly site will remain active for 90-days, and then will be unpublished”	Same
December 26, 2026	Account access ends; customers are told to download data and transfer domains before this date	Same
No date	Weebly continues, unchanged, with new users routed to Square Online	United States, UK, Canada, Australia and most of Western Europe (not on the list)

I want to be precise here, because a lot of what is written about this online is not. The 67 countries are listed on the notice: Taiwan, Singapore, Malaysia, the UAE, Thailand, South Korea, Colombia, Chile, Turkey, Saudi Arabia, Vietnam, Egypt, Peru, Nigeria, Kenya, Ukraine, Russia and so on. The United States is not on it. SoSlanty, in Ohio, was not going to be switched off on September 27. If you are a US, UK, Canadian or Australian merchant reading a headline that says “Weebly is shutting down,” the accurate version is “Weebly is shutting down for some people and has stopped moving for everyone.”

That second clause is the real reason to leave, and it was the reason SoSlanty left. The owner’s brief did not mention the shutdown at all. It asked for an automated Merchant Center feed instead of a spreadsheet she updated by hand, a mega menu, a sticky header, product filters that were not stuck on the left, a stock counter, a second image on hover, quick view, reviews she could actually manage, and, in capitals, “DO NOT WANT INFINITE SCROLL.” Every one of those is table stakes on a modern Shopify theme and none of them was coming to Weebly. When a platform has had no meaningful ecommerce feature in years and its owner is actively closing it in a third of the world, the question is not whether you will move but whether you move on your own schedule or on theirs.

There are also structural limits that no amount of patience fixes. Square Online processes payments through Square, so there is no Shop Pay, no PayPal at checkout without workarounds, and no negotiating the rate. The variant model is shallow: options exist, but multi-location stock, wholesale pricing and bundles do not. The app marketplace is short, and the tools growing brands lean on are built for Shopify first. Product URLs are routed by numeric ID, metadata control is limited and the templates are slow, all of which make ranking harder than it should be. None of that is a criticism of Weebly as it was in 2015. It is a description of a product that has been left where it was.

One practical point before moving on, because it changes the risk profile of the whole project. Find out early where the domain and the email actually live. SoSlanty’s domain was registered and its DNS hosted at a third-party registrar, not at Weebly, which meant the cutover was a record change rather than a domain transfer with a 60-day ICANN lock hanging over it. Her Google Workspace, on the other hand, was billed through Weebly, and the self-service transfer option Square describes was simply not present in her account; that took a support conversation and was only confirmed complete on September 17, the day Weebly was cancelled. Had we cancelled Weebly a week earlier, the business email would have gone with it. Square’s own notice tells wind-down customers to transfer domains before December 26 and to contact support about Workspace; treat both as week-one tasks, not launch-week ones.

Where Weebly’s Data Actually Lives: Square Online, a Thin Export and a 1,000-URL Sitemap

The first surprise on any Weebly migration is that the data is not where the owner thinks it is. Logging into the Weebly account lands you in the Square dashboard at squareup.com, where Items, Inventory and Reports all appear empty, because a store built on Weebly never populated the Square side. The store data lives under Online, behind a site switcher, at what is now square.online. Items are at Items › Site Items; orders are exported from Orders › Export Orders; and the product export is not downloaded but emailed to you a few minutes later. We now put this in the first email to a Weebly client, because finding it took longer than it should have.

The second surprise is how thin the export is. Square’s item export for SoSlanty was one CSV of 1,663 rows across 16 columns: title, a single image URL, categories joined by commas, a taxable flag, a SKU field that actually held an internal ID, one row per variant with no option names, price, inventory and weight in pounds. That is it. Compare it with what Shopify needs to build a product, and the gaps are the migration.

Data	In the Square Online export?	Where it came from instead
Title, price, stock, weight, categories, taxable flag	Yes	The export, transformed and used as the reconciliation baseline
Product descriptions	No	Harvested from the live product pages by script
Product URLs (for redirects)	No	Sitemap, the old Google feed, then a probe of every product ID on the live site
Option names (Size, Color)	No; variant rows exist but the option is blank	Rebuilt from the variant rows and the live pages
Images beyond the first	No	Harvested from the live product pages
Sale prices	No	Not carried; compare-at pricing decided later with the owner
Order history	Separate orders export, 47 columns, one row per line item	Our importer, through the Admin API, for the bulk and for the final delta
Tracking numbers on shipped orders	No	Not available anywhere; noted to the owner
Customers and marketing consent	Square Customer Directory export	Our importer; consent taken from the directory, not the orders
Product reviews	No export exists (confirmed with Weebly support)	The owner retyped 113 reviews into a template we supplied
Store (home page) reviews	No export; rendered by a third-party widget	Rescued from the rendered page, 110 of them
Theme and design	No	Rebuilt on Shopify’s Horizon theme

The third surprise sits in the sitemap. Weebly’s sitemap.xml lists at most 1,000 products. SoSlanty had 1,658. Anyone building a redirect map from the sitemap, which is what most automated tools and most agencies do, would launch with several hundred product URLs unaccounted for and would find out when the 404s started. Weebly product URLs also do not follow the pattern you would guess from the title: they are routed by numeric ID in the form /store/p1234/some-slug.html, the slug is whatever the page was called when it was first created, and around forty of SoSlanty’s were in mixed case. Title-based guessing would have missed every one. We deal with this properly in the redirects section; the point here is that a Weebly store’s URL inventory is not something you can export. You have to go and get it while the site is still alive.

There is one honest conclusion from all this. Shopify’s own migration documentation lists eleven steps and starts with “Import your store’s content and data,” which assumes you have an export worth importing. On Weebly you have a fifth of one. The rest is assembled from the live site and from the owner’s own records. Budget for that, and the project goes smoothly. Assume a clean CSV, and you will be writing apology emails.

Running the Migration on Our Own AI-Driven Technology: Products, Customers, Orders and the Checks Between

There are off-the-shelf migration apps that list Weebly as a source. We started down that road and stopped, because a generic connector cannot know what a particular platform’s data means, and on Weebly the meaning is most of the work. Instead the whole migration was run on the AI-driven migration technology we have built in-house: an agentic toolchain that writes the import scripts for a specific source platform, runs them against Shopify’s Admin API, reads the results back, and reconciles every entity against the source exports in the same working session. Every script is reviewed by a person before it runs, every write is idempotent so it can be re-run safely, and every bulk change keeps a restore path. We described the approach in July in what happened when we let an AI agent run a Shopify Plus migration; SoSlanty is the small-store version of the same method.

The order of operations matters on Shopify. Products go first, because customers and orders reference them; then customers, so that orders can be attached to the right accounts; then orders, written with their original numbers so that a customer’s order 2847 from 2021 is still order 2847. Here is what each entity needed, where it came from, and how it was checked.

Entity	Source	How it was moved	Verified against
Products, variants, prices, stock, weight	Square Online items export (1,663 rows)	Transformed into Shopify products with rebuilt option structure; one hidden product recreated by hand	Export row by row: 1,658 products, 1,666 variants, 8,973 units matched exactly
Descriptions and additional images	Live Weebly product pages	Harvested by script before the site went dark; 1,511 descriptions cleaned of empty spacer paragraphs	Spot checks plus a count of products with fewer than two images
Taxable flags	Per-row flag in the export	Applied on import and re-applied from the final export on go-live day	1,599 taxable and 63 exempt, matching Weebly exactly
Categories → collections	Export categories plus the old category pages	44 collections with the old meta titles and descriptions preserved; 31 retitled, 2 split, 29 regrouped by item type	Membership counts against the export
Vendor and product type	Not in the export	Vendors backfilled from titles and the owner’s knowledge (32); product types seeded by a title classifier (56 types)	Owner review; 8 products still awaiting a real brand
Customers and consent	Square Customer Directory export	Accounts created; marketing consent set only where the directory said so	Directory row count; consent flags sampled
Orders with line items and status	Square orders export (3,412 orders, 47 columns)	Written through the Admin API with original order numbers; pending orders modeled as authorized	Order-number diff: 3,386 in the bulk run, 26 in the final delta, none missing
Gift card	A placeholder product on Weebly	Real Shopify gift card product created; old placeholder archived and its URL redirected	Test purchase of a gift card
Reviews	No export exists	Handled separately (see the storefront section)	Counts on the home page and product pages

Two of those rows deserve a paragraph each, because they are where a migration that looks finished is not. The first is tax. Square Online records a taxable flag per variant, and the export carries it, but nothing on the Shopify side enforces it: a product created without the flag simply does not collect sales tax, and the admin gives no warning. On go-live day we diffed the final export against the store and pushed the flag on every variant; 1,599 came out taxable and 63 exempt, exactly as Weebly had them, and the store collected tax correctly from its first Shopify order. Had that step been skipped, roughly 96% of the catalog would have under-collected until somebody noticed on a tax return.

The second is the fields Weebly never had. There is no vendor on a Square Online item and no product type, only a category. Shopify’s filters, its product taxonomy and, later, Google Shopping all lean on those two fields, so we backfilled 32 vendors and 56 product types by script from the titles and from the owner’s knowledge of her own suppliers. Eight products still carry a placeholder while she confirms the brand. That is the honest state of it, and it is the kind of detail the admin’s product count will never tell you.

Why does the tooling matter to a reader who is not going to write scripts? Because the alternative is trusting a summary screen. A connector reports matching counts, and matching counts were true here too: 1,658 products, 3,412 orders, 3,213 customers. The store still would not have been fit to launch without the tax pass, the vendor backfill, the description cleanup and the URL harvest, none of which any counter reports. Running the migration on our own technology meant the same session that imported the data could also interrogate it, and that is the difference between a migration that is complete and one that merely looks complete.

The Redirect Map: 1,758 Redirects for URLs Nobody Could Guess

Redirects are where SEO equity lives or dies in a migration, and Weebly makes them harder than any platform I have worked with. Recall the shape of the problem: URLs are routed by numeric product ID, the slug after the ID is frozen at creation time, the sitemap stops at 1,000 products, and the export contains no URLs at all. A redirect map built by slugging product titles, which is what an automated tool generates, is wrong on Weebly by construction.

We built the map in two passes. The first, on July 28, took the union of every product URL in the sitemap and every product link in the owner’s Google Shopping sheet, which between them yielded 1,553 product paths, plus 44 canonical category URLs in the /store/cN/ form, 46 “pretty” category URLs ending in .html, and the static pages (about, contact, policies, the reviews page, the gift card page). That produced 1,651 redirects. It also left roughly 105 products with no discoverable URL, all of them beyond the sitemap cap and absent from the feed.

The second pass ran on go-live day, while Weebly was still serving. Every Weebly product answers at /store/p<id>/anything.html and responds with a redirect to its canonical slug, so a script requested each of the missing IDs against the live site and recorded where it landed. That harvested the exact canonical URL for every product, including the mixed-case slugs and a handful of products that had been renamed and still answered on both the old and the new slug. It added 106 redirects and brought the total to 1,758, covering 1,657 of 1,658 products. The one exception was a product hidden on Weebly with no public URL to redirect from. The complete ID-to-URL inventory is saved in the project as a JSON file, because after Weebly goes dark it cannot be recreated.

Source of old URLs	URLs found	Redirects created	What it missed
Weebly sitemap.xml	1,000 products (hard cap)	Included in pass 1	Everything beyond product 1,000
Owner’s Google Shopping sheet links	Additional products up to 1,553 combined	Included in pass 1	Products never added to the feed
Category URLs (canonical and pretty)	44 + 46	90	Nothing; both forms covered
Static pages	8	8	Nothing
Live-site probe of every product ID	106 more canonical product URLs	106	One hidden product with no public URL
Total		1,758	1,657 of 1,658 products covered

Then every redirect was tested, not by reading the list in the admin but by requesting the old path against the new store and confirming a 301 to a page that returned 200. The store was password-protected before launch, so the test script first authenticated through the password wall with a cookie jar, then walked the list. Two hours of script time bought a launch with no dead product links, on a catalog where the median item costs thirteen dollars and no customer would have bothered to report one.

After launch, redirects are not finished; they are monitored. Google keeps crawling old URLs for months, and the Search Console coverage report is where the stragglers show up. That 30-day watch is part of the post-migration SEO evaluation we run on every store, and it is why I am comfortable saying SEO equity was preserved rather than hoping it was.

Rebuilding the Storefront in Horizon: Menus, Sayings and 110 Rescued Reviews

Design does not migrate. It never has, from any platform, and Weebly is no exception: the theme, page layouts, banners and widgets all stay behind. We rebuilt SoSlanty on Horizon, Shopify’s current flagship free theme, version 4.1.3 at the time, which we use as the default on client builds because its block-based sections cover most of what a small retailer needs without paid theme licenses or apps. The owner’s feature list from the brief maps onto Horizon’s built-in settings almost line by line: sticky header, hover image, quick view, stock counter, filters, and, as requested, no infinite scroll.

Some of the work was ordinary theme configuration. Some of it was recovering things the old site had that no export contained, and that is the part worth describing, because it is what makes a migrated store feel like the same business rather than a template with the same products.

What the old Weebly site had	How it came across	Detail
Six-column drop-down menu	Rebuilt from the owner’s exact tree	Text-only mega menu, wrapping horizontally, non-navigating parents, wavy hover underline; the whole tree pushed in one API call
Category “sayings” (e.g. “Celebrate Your Sparkle”)	11 recovered, 35 written with the owner	Sayings lived in Weebly page banners, not in the category data; one was typed letter-spaced. The owner filled a spreadsheet for the gaps, ten in her own voice
Animated category headlines	Recreated	Per-character drop-and-wobble animation in the Chewy font, inlined into the theme; auto-shrinks for long quotes
Rotating announcement bar	Recreated, plus a fix	Horizon stops rotating on mouse-over; a small script resumes it
Breadcrumbs	Added	Horizon ships without them
110 home-page reviews, 4.93 average	Rescued	Rendered by a third-party widget with no export; captured from the page and imported to Judge.me as store reviews
Product reviews	113 typed by the owner, 112 imported after de-duplication	No source export; Judge.me’s importer runs in a cross-origin iframe, which took some persuading
Blue and red logo palette	Applied theme-wide	Color schemes, buttons, wavy dividers and five color banners
Shop All page	Rebuilt	Horizon caps the collection list at 20 items; a core patch lifts it to 60, and is on the update checklist
HERstory page (a 15-page PDF of women’s history products)	Rebuilt as a page	82 rows extracted from the PDF; 38 product links repointed to Shopify

The reviews are the part I would underline for any Weebly owner. Weebly support confirmed there is no review export. The 110 store reviews existed only as HTML on the home page, so we captured them from the rendered site and imported them into Judge.me, which needed a placeholder email per reviewer and could only approximate the dates. The product reviews did not exist anywhere but in the owner’s memory and inbox, so she typed 113 of them into a spreadsheet template we supplied, and 112 survived de-duplication and went in on go-live day against 89 products. Nine years of social proof would otherwise have vanished the day the old site went dark. It took the owner an evening and us a script. It is the best-value work in the project.

Horizon has its own quirks, which we document as we go: a 500-character limit on per-section custom CSS, a theme editor that silently drops an invalid section file, a CLI that will not push certain file types to a live theme, and app embeds that live inside the theme settings file, which means a careless settings upload can erase a review widget from every page. We did that once, on July 27, and restored it the same day. I mention it because a case study that only lists successes is not one you should trust.

Six days before launch the owner wrote “only the home page looks good,” with screenshots attached. We walked through the store together, page by page, on the August 31 call, and she sent the go-live email two days later with the words “I am excited!” I include that not for sentiment but because expectation management around preview links and half-styled pages is a real deliverable on these projects, and it is cheaper to do on a call than by email.

Go-Live Day, Step by Step

Go-live was Wednesday, September 2, 2026. The store had been built in a free Shopify development store, which is one of the practical benefits of working with a Shopify Plus Partner: SoSlanty paid Shopify nothing until this day. The plan was a delta sync from the last Weebly exports, a final reconciliation, the plan upgrade, the DNS change and a set of checks that we run in the same order every time. Here is how the day went, in order.

Order	Step	Result
1	Owner exports final products and orders from Square Online	Files arrived about 2:15 pm Eastern; the orders export held 26 orders and 26 customers placed since the July import
2	Taxable-flag diff and push	1,599 variants to taxable, 63 left non-taxable, matching Weebly
3	Inventory sync from the final export	27 variants adjusted
4	Delta orders and customers imported with original order numbers	26 + 26, tagged for traceability; one pending order tracked and fulfilled the next day
5	Product-ID probe against the live Weebly site	106 missing product URLs found; redirects at 1,758
6	112 product reviews imported to Judge.me	89 products now carry reviews
7	Free-shipping rule added ($0 at $95 and over) via the delivery profile API	Verified on a live $101.78 test cart
8	Store moved to a paid Basic plan; owner adds the business address	Which fixed a problem we had not yet noticed (below)
9	Order-confirmation email branded; DNS pointed at Shopify	Old-site analytics show one session the next day, then nothing
10	“Okay, we’re live on Shopify,” 4:38 pm Eastern	Total build time: 44 days from kickoff

The problem we had not yet noticed is the one I would warn every Shopify migrator about. During testing, carts under $95 showed “Shipping not available” at checkout. Everything in the shipping settings looked right: a domestic zone covering every US state, USPS calculated rates with a $1.75 handling fee, the free-shipping rule above $95. The cause was that the store’s only location had no address. Calculated carrier rates need an origin, and a development store starts with a location that has none. The moment the owner entered the business address, rates appeared. We now verify it with two HTTP requests rather than a browser: add an item to the cart, then request shipping rates for a test ZIP. On September 17 a $7 cart returned USPS Ground Advantage at $7.33 and Priority Mail at $9.76, both carrying the handling markup. That pair of requests is on our launch checklist for good.

The other piece of launch-day plumbing worth recording is access. The moment a store leaves development status and becomes a paid store, an agency working as a collaborator can no longer create the kind of private app credentials it used during the build. The orders importer needed a proper offline API token. Shopify’s documented route for a collaborator is an authorization-code grant against an app of your own, and it has two prerequisites that produce an unhelpful “Oops, something went wrong” if you miss them: the app configuration must list a redirect URL, and it must use the legacy install flow. We had solved this on a previous project, so it cost us little time, but it is the sort of thing that stops an unprepared launch cold. Shopify’s Basic plan, incidentally, is where SoSlanty landed: a one-person store does not need staff accounts, and the plan can be raised later without touching anything else.

Two things happened after launch that belong in the timeline. On September 5 the owner asked about a Shopify prompt to authenticate her sending domain, and we found the domain had no SPF record at all; SPF and DMARC records were added at her registrar, and Shopify’s sender authentication was confirmed a week later. And on September 17, once the last pending Weebly order had been confirmed settled and the Google cutover (next section) was done, the owner cancelled Weebly. The old site had run for fifteen days after launch, which is exactly as long as it was needed and not a day more.

The two weeks between launch and the Weebly cancellation were not idle. A migration produces a punch list the moment real customers arrive, and SoSlanty’s is a fair sample of what to expect. The owner asked what registered customers could see and do, so we wrote it up; she asked for a sign-up popup with a coupon, which Shopify now does natively without an app; a product filter on collection pages was clipping any list longer than eight values with no scroll cue, fixed with a few lines of CSS; the gift card product went live on September 12 after Shopify’s gift-card feature had been switched on from the admin, which cannot be done through the API; the HERstory page, a fifteen-page PDF catalog of women’s history products, was rebuilt as a page with its 38 product links repointed; and because no native Shopify report groups inventory by collection or vendor, we produced a spreadsheet that does, for reordering. On September 13 a classifier seeded 56 product types across 1,685 products so the filters had something to work with. None of this is glamorous. All of it is the difference between a store that launched and a store that runs.

Keeping the Store’s Google Shopping History

This is the section I most wanted to write, because it is where most Weebly to Shopify migrations quietly lose the most value, and nobody notices for a month. SoSlanty’s Google Shopping campaign was small, ten dollars a day, but it had been running long enough for its products to carry real performance history in Merchant Center, and that history is attached to the product IDs. Merchant Center identifies a product by three things together: content language, feed label and offer ID. Google’s documentation for the id attribute is explicit that the ID should not change between updates; a changed ID is treated as a new product, and the history stays behind with the old one. SoSlanty’s IDs were BB1 through BB1707, from the owner’s hand-maintained sheet.

Shopify’s native Google and YouTube channel writes its own IDs in the form shopify_US_<productId>_<variantId>. There is no setting to change that. Connecting the channel would have submitted 1,685 brand-new products and let the old ones expire, and the campaign would have started from zero. The owner had specifically asked for an automated feed to replace the sheet; the obvious automated feed would have thrown away the thing she was trying to protect.

The solution has three parts, and the order matters. First, set the store’s SKUs to the old feed IDs, so the join is durable and visible in the admin. A script matched 1,476 products to their BB IDs by the Weebly product ID embedded in each sheet link, which also removed 29 earlier title-based mismatches, filled 31 blank SKUs, repaired 16 barcodes and cleared 36 invalid ones. Second, install a feed app that can submit SKU as the product ID. We used Simprosys Google Shopping Feed, which after its trial costs $13.99 a month and, usefully, reads Google attributes from a standard metafield namespace, so the same script wrote gender, age group, color and size for 1,530 products and the app picked them up without mapping. Third, keep the old feed alive until the new items are visibly serving, then remove it.

That third step is where the surprise lived. A Merchant Center data source’s feed label is fixed at creation. The app submitted under one label; the old sheets used another; and because feed label is part of a product’s identity, Merchant Center treated them as separate products running in parallel. The Shopping campaign had to be switched to use all available feeds, a one-way change that permanently removes the older country-of-sale selector. Within a day, items that existed only in the new feed were taking impressions and clicks, which was the proof we needed, and on September 18 the old sheet sources were deleted. Total elapsed: about a day and a half from the first SKU change to a fully automated feed with the history intact.

Google asset	The risk in a Weebly to Shopify migration	What we did
Merchant Center products	New product IDs reset performance history; the old sheet’s images 404 the day the site moves (295 of 1,370 prices were already wrong)	SKU set to the legacy ID; feed app submits SKU as ID; old feed kept until new items served, then removed
Merchant Center website claim	Changing the store URL drops the claim and disapproves the catalog	Verified the domain in Search Console first; the URL change then claimed automatically
Search Console	“There are no sites listed” looked like the property had never existed	Ownership history showed a 2023 verification and an unused token; re-adding the owner restored access instantly
Google Ads conversion tracking	A setup wizard creates new conversion actions, orphaning the ones with history; GA4-imported and tag-based purchase actions can both be Primary and double count	Reused the existing purchase action; demoted the GA4 import to secondary
GA4	The native channel’s Analytics step is locked behind Merchant Center, and its dropdowns defaulted to other clients’ accounts on our agency login	A Shopify custom pixel sending ecommerce events to the existing GA4 property
Structured data	Theme and review app each emitted a Product node with the same ID; Google evaluated them separately and flagged missing fields	Merged the rating into the theme’s node; switched off the app’s duplicate
Apparel attributes	53 products with no color were “eligible (limited)” and in zero campaigns	Color set to multicolor where the product genuinely is; the count went to zero

The GA4 pixel took three iterations and taught us two things I have not seen written down anywhere. The first is that Google Ads frequently issues the same measurement ID as the store’s GA4 stream, so any app that loads the Google tag for Ads is already sending a GA4 page view; add your own and every page counts twice. The fix is to send only the shopping events from the pixel and address them explicitly to the GA4 stream. The second is that inside Shopify’s custom pixel sandbox, the document cookie is a cached copy that refreshes asynchronously, so a second Google tag instance can mint its own client ID on a visitor’s first page and split one visitor into two. The working version waits for the other tag’s cookie before loading. You cannot see any of this from the browser’s network panel, because pixel traffic runs in a cross-origin frame; we verified it with a debugging-protocol capture. Four runs, one client ID, one session ID.

By September 18 Merchant Center showed 1.69 thousand products total, 1.66 thousand approved, zero limited and 14 not approved, with approvals up 239 week on week as the feed settled. Before the cutover, the old sheet-based feed had shown clicks down 35.6% over 28 days, as its Weebly-hosted images died. We have written the general version of this up as a Google Shopping feed migration service, because it applies to every replatform, not just Weebly.

The Numbers: What Moved, What Was Verified and What We Did Not Measure

A case study is only as good as its honesty about measurement. Here is what we can state as fact from the project files, followed by what we cannot yet.

Measure	Figure	How it was verified
Products live	1,685 active, 1,693 variants	Admin counts against the final Square Online export; 100% assigned to Shopify’s product taxonomy
Orders	3,412 with original order numbers	3,386 bulk plus 26 delta, diffed on order number
Customers	3,213	3,187 bulk plus 26 delta
Redirects	1,758; 1,657 of 1,658 products covered	Every redirect requested through the storefront; 301 to a 200
Sales tax flags	1,599 taxable, 63 exempt	Diff against the export; matches Weebly’s own flags
Inventory	8,973 units at kickoff, re-synced at launch	Matched the owner’s export to the unit; 27 variants adjusted on launch day
Reviews	108 store reviews, 112 product reviews on 89 products	Imported to Judge.me; visible on the home page and product pages
Checkout shipping	$7.33 / $9.76 on a $7 cart; free at $95+	Live cart tests on September 2 and 17
Merchant Center	1,660 of 1,690 approved, 0 limited, 14 not approved	Merchant Center dashboard, September 18
Feed continuity	Legacy IDs preserved on 1,476 products; new feed serving within a day	Impressions and clicks recorded on items that exist only in the new feed
GA4	One client ID, one session per visitor	Four instrumented capture runs
Sitemap	Submitted and fetched with 0 errors	Search Console
Timeline	44 days kickoff to live; 62 from first inquiry	Email and project log dates: inquiry July 2, kickoff July 20, live September 2

What we have not measured, because it is too early or because the data did not exist before: conversion rate before and after, since the Weebly analytics were bare purchase events and the new store’s first tracked purchase had not yet landed when this was written; revenue lift, for the same reason; organic traffic after launch, which needs at least a full month of Search Console data to say anything meaningful; Core Web Vitals, which we will report once the field data exists; and Ads performance after the feed switch beyond the recovering click trend. I would rather publish those in a follow-up than estimate them now, and if you have read migration case studies that quote a 40% conversion lift two weeks after launch, you know why.

What I can say qualitatively is that during the 44-day build the old store took 26 orders, roughly four a week, and the new store was taking orders the same week. There was no gap in trading. The owner had a working store the whole time, paid Shopify nothing until the day it was needed, and lost none of her products, orders, customers, reviews or Google history. That is the result the project was scoped to deliver.

A result that does not show up in a table is whether the owner can run the store without us. Before launch she received a written owner’s guide to running the store day to day, alongside a site-update document that answered each of her feedback items in turn. After the Google cutover she received a runbook page for the feed, Merchant Center and Ads, with the things that must not be touched marked as such. She also received two spreadsheets listing 34 invalid barcodes and 17 barcodes shared across 38 products, all of which pre-dated the migration in her own feed and are hers to correct with her suppliers. A migration that leaves the owner dependent on the agency for routine work has not finished; it has just changed who the store depends on.

What This Ecommerce Migration Case Study Teaches About Any Weebly to Shopify Move

The value of an ecommerce migration case study is in what transfers to the next project. Almost everything above does. Here are the lessons, in the order you will meet them.

1. The data is in Square Online, not Square. Go to Online › Items › Site Items and Orders › Export Orders. The product export arrives by email. Do this before anything else, because those exports are the baseline every later check is measured against.
2. Treat the export as a checklist, not a source. Sixteen columns, one image, no descriptions, no URLs, no option names. The rest comes from the live site and from the owner. Scope the project on that basis.
3. Reconcile against the source, not the destination. Matching counts in the Shopify admin prove nothing. Taxable flags, vendor, product type and description cleanliness can all be wrong at matching counts. Diff field by field, in the same session that did the import.
4. Apply the tax flags deliberately. Shopify will not warn you about a product that does not collect tax. Push Weebly’s per-variant flags on import and again from the final export before the first real order.
5. Harvest the URL inventory while the site is alive. The sitemap caps at 1,000, slugs are frozen at creation and sometimes mixed case, and the export has no URLs. Probe every product ID and save the map; it cannot be recreated after the site goes dark.
6. Test redirects through the storefront. A 301 in a list is a promise. A 301 that lands on a 200, requested through the password wall, is a fact.
7. Rescue reviews early. There is no export. Home-page reviews can be captured from the page; product reviews come from the owner. Give her the template in week one, not week five.
8. Look for the sayings. On Weebly, the personality of a category often lives in a page banner rather than in the category record. Sweep the old category pages before they disappear.
9. Give the store an address before you test checkout. Calculated rates with no origin location produce “Shipping not available,” and nothing in the shipping settings tells you why.
10. Plan the API access change at go-live. The credentials that worked in the development store stop working the moment the plan is paid. Have the authorization-code route ready.
11. Never connect the native Google channel if history matters. It rewrites product IDs. Use SKU as the ID through a feed app, set SKUs to the legacy IDs, and run the old feed in parallel until new items serve.
12. Get a verified Search Console owner in place before changing Merchant Center’s URL. Read the ownership history before concluding a property was never verified; an unused token is a five-second fix.
13. One Google tag, addressed explicitly. Ads and GA4 often share a measurement ID. Send shopping events only, wait for the existing cookie, and verify outside the browser.
14. Cancel the old platform last. Weebly holds the pending orders, the live pages you may still need to harvest from, and possibly the Google Workspace billing. Confirm all three are clear before you close it.

If your store is on Weebly or Square Online and any of this sounds familiar, our migration guide for the platform turns these lessons into a checklist, and the platform directory covers the other 88 carts we move from.

A Quick Overview of Ruby Digital Agency (and Why We Care)

For readers meeting us for the first time: Ruby Digital Agency is a Shopify Plus Partner that does one thing on purpose. We move brands onto Shopify and Shopify Plus and make the move pay for itself. Our clients arrive from WooCommerce, BigCommerce, Magento and Adobe Commerce, OpenCart, Shift4Shop, PrestaShop, Weebly and Square Online, Wix, and custom PHP builds. Smaller, owner-run stores are usually live in six to eight weeks; a standard migration for a mid-market brand doing roughly $2M to $20M a year runs 8 to 12 weeks; and an enterprise replatform to Shopify Plus, typically from Magento, Adobe Commerce or a custom stack, runs three to six months. All of them are delivered with zero downtime and with SEO equity preserved through a complete, tested 301 redirect map, and in all of them the specialists who architect the project are the ones who deliver it. We are a boutique team, with no account-manager layer between you and the people doing the work. Because we build in free Shopify development stores, you pay Shopify nothing until the day you go live.

The economics are why brands call us. Across our migrations, clients recover the investment in under 18 months, see conversion rates about 15% higher on average, and cut total cost of ownership by around 29%; infrastructure costs alone fall by as much as 80% when a self-hosted stack goes away.

Service	Who it is for	Typical timeline
Standard Shopify Migration	Mid-market brands, $2M–$20M/yr	8–12 weeks
Shopify Plus Replatforming	Enterprise brands, $20M+/yr, from Magento, Adobe Commerce or custom builds	3–6 months
Small-Store Migration (Weebly, Square Online, Wix, Squarespace, Etsy)	Owner-run stores with modest catalogs	6–8 weeks
Google Shopping Feed Migration	Any store with Merchant Center history to protect	Included, or scoped on its own
Ecommerce Website Audits	Any platform: Shopify, WooCommerce, BigCommerce, Magento, custom	Fixed-scope engagement
RDAI SEO and AEO	Brands that need to be found in Google and cited by AI assistants	Ongoing
Shopify POS Implementation	Retailers uniting stores and online	Scoped per project

Some of the work is public. Midwood Guitar Studio came from BigCommerce with POS; D.T. Medical Devices from WooCommerce in Germany; Earache Records and The Gadget Guys from OpenCart. The full list is on our case studies page. In July we published what happened when we let an AI agent run a Shopify Plus migration, with the data.

Why does a small Weebly store get the same treatment as a Plus replatform? Because the failure modes are identical and the owner has less margin for them. A dead redirect or an under-collected tax flag hurts a one-person gift shop more than it hurts a brand with a finance team. The scripts that ran SoSlanty’s migration, the import, the reconciliation checks, the URL probe, the feed-ID preservation, were produced and executed by our in-house AI-driven migration technology and verified by a person who is accountable for them. That is how all our work is done now, and it is what makes a small-store migration this thorough possible at all.

A Ten-Step Plan for Your Own Weebly to Shopify Migration

If you run a Weebly or Square Online store, this is the order we would follow, whether you do it with us, with another agency, or on your own with a migration tool. Each step exists because skipping it cost somebody something in the project above.

1. Export everything from Square Online today. Items, orders with no date filter, and the Customer Directory. Save the files somewhere permanent. If you are in one of the 67 wind-down countries you have until December 26, 2026 to log in at all; if you are not, do it anyway, because these exports are your baseline.
2. Harvest your URL inventory while the site is live. Pull the sitemap, then request every product ID you own and record where it lands. If you have more than 1,000 products the sitemap alone will strand the rest. Keep the ID-to-URL map with the exports.
3. Open a free Shopify development store through a partner, or a trial if you are going alone. Build there. You should pay Shopify nothing until the day you switch.
4. Import products, then customers, then orders, in that order. Products from the export, with descriptions and images harvested from the live pages; customers from the Customer Directory; orders through the Admin API with their original numbers preserved. Whether you script it, as our migration technology does, or use an app, read the next step before you trust any of it.
5. Reconcile against your exports, field by field. Taxable flags, inventory, prices, vendor, product type, descriptions, option names. Expect vendor and product type to be empty, because Weebly never had them. Fix them in bulk before anything else is built on top.
6. Rebuild the storefront on Horizon and recover what the export forgot. Menu tree, category sayings, banners, policies, the reviews page. Capture home-page reviews from the rendered site; give the owner a spreadsheet for product reviews in week one.
7. Build the redirect map from the harvested inventory, then test it through the storefront. Product URLs, both forms of category URL, static pages, the gift card page. Confirm each old path returns a 301 to a page that returns 200.
8. Set up shipping, taxes, payments and the store address, and test a cheap cart and an expensive one. The address is not optional; calculated rates need an origin. Verify the free-shipping threshold and the calculated rates on a real ZIP.
9. Prepare the Google cutover before you move DNS. Set SKUs to your legacy Merchant Center IDs, install a feed app that submits SKU as ID, confirm a verified Search Console owner sits among the Merchant Center users, and decide which Ads conversion action holds your history. Do not connect the native Google channel.
10. Launch with a delta sync, then run the Google switch and a 30-day watch. Final exports, diff, import the new orders and customers, move DNS, submit the sitemap. Change Merchant Center’s URL, watch the new feed serve, delete the old one, and read the Search Console coverage report weekly for a month. Cancel Weebly last.

What Happens If You Do Nothing

If you are in one of the 67 countries on Square’s list, doing nothing has a date on it. Your site stops being editable, then stops being visible on September 27, then your account closes on December 26, 2026. Your domain, if Weebly registered it, needs to be transferred with an EPP code before then. Your product reviews, which have no export, are gone with the site. The Google Shopping history attached to your product IDs expires as the images 404. None of that is recoverable afterwards, and the ID-to-URL map that makes clean redirects possible cannot be rebuilt once the site is dark. This is not a “migrate eventually” situation; it is a “download everything this week” one.

If you are in the United States or another market Weebly still serves, the cost of doing nothing is quieter and, over a year, larger. You keep paying for a platform that does not add features, cannot take Shop Pay, limits your variant and inventory model, and gives you ID-based URLs that make ranking harder than it should be. Every month you stay, your Google feed remains a spreadsheet someone updates by hand, your reviews stay locked in a widget, and the eventual move gets harder because there is more history to protect and less warning when the market list grows. Square has already announced the end of the mobile app and directs every new customer elsewhere. The pattern is not ambiguous.

The practical middle ground, if you are not ready to move, is to do steps one and two of the plan above now: export everything and harvest your URL inventory. It costs an afternoon, it takes nothing away from the running store, and it turns a forced migration later into an orderly one. SoSlanty moved on her own schedule, with the old store running until fifteen days after the new one launched. That was possible because the data had been secured first. The stores that move badly are the ones that start exporting after the notice arrives.

Frequently Asked Questions

Final Thoughts

The owner’s two questions from the brief deserve their real answers now. How many Weebly to Shopify migrations have we executed? Enough to know where the bodies are, and the count goes up every month as Square’s market list gets longer. Any particular quirks to look out for because Weebly is so old? Yes: the data lives in Square Online, the export is a fifth of a catalog, the URLs cannot be guessed, the reviews cannot be exported, the tax flags arrive wrong, and the Google feed will reset itself if you let the obvious tool touch it. None of those is hard once you know it is there. All of them are expensive if you find out after launch.

What I would tell a Weebly or Square Online owner in one sentence is this: the platform has already decided to leave you, so decide when you leave it. If you are in a wind-down country, that decision has a date. If you are in the United States, the date is whenever you would rather spend a fixed six weeks moving on your terms than an unknown number of months waiting for the list to grow. Either way, export everything and harvest your URLs this week, because those are the two things that cannot be done later.

SoSlanty is live, taking orders, with nine years of history intact and a Google feed that updates itself. That is what a migration is supposed to look like: not dramatic, fully checked, and finished on the day it was promised. We will publish the traffic and conversion numbers when there are enough of them to mean something. In the meantime the holidays are closer than they look, and somebody on your list needs a pair of socks that says exactly what you think of them. Start there.

Related Reading

• Weebly and Square Online to Shopify Migration Guide (2026)
• Google Shopping Feed Migration: Keep Your Merchant Center History
• The Ecommerce AI Knowledge Base: How One MCP Server Powers Web Chat, Voice Support and AI Search in 2026
• Case Study: Midwood Guitar Studio from BigCommerce to Shopify with POS
• We Let an AI Agent Run a Shopify Plus Migration — Here’s the Data
• All Ruby Digital Agency Migration Case Studies
• Shopify Migration Services: Every Platform We Move From
• More Shopify Guides on Our Blog

Lance Allison
Founder & CEO, Ruby Digital Agency

Lance Allison is a Shopify Plus Partner and eCommerce migration specialist based in Salt Lake City, Utah. He has helped hundreds of merchants move to Shopify and Shopify Plus, and specializes in complex, data-heavy platform migrations, including the small-platform moves from Weebly, Square Online, Wix and Squarespace where the data is thin and the history is worth protecting.