Shopify Integration FAQ: Connecting, Imports, Variants and Catalogue Scale
How Klip Kanvas works with Shopify — connecting a store, what product data gets imported, variants and collections, catalogue-scale generation, refreshes, failures and limits.
Shopify is the fastest way to load a catalogue into Klip Kanvas, but it is an import, not a second admin. This hub covers connecting, what actually arrives, variants, collections, scale, refreshes, common failures and the limits to plan around.
01
Connecting Your Store
1.How do I connect my Shopify store?
Two paths. Paste a product or collection URL and we pull that page immediately — that is enough for a first ad in under 10 minutes. For a working catalogue inside the workspace, connect the store so products are selectable without pasting SKU by SKU. You need access to the store, not a staff screenshot of the admin. Password-protected storefronts, stores in a paused plan, and checkout-only custom domains are the usual blockers. Connecting does not install a theme, does not alter products, and does not publish ads to Shopify. It only reads catalogue data into your workspace.
#Connecting Your Store
2.Do I need a Shopify app install, or is a URL enough?
A URL is enough to generate. A store connection is what you want if you will make ads for more than a handful of SKUs, because you can pick products from a list, keep images attached, and refresh later. If you are evaluating the free 50 credits, paste one product URL and do not wait on IT. Agencies should connect inside the client workspace, not their house workspace, so access ends with the contract. If your merchant refuses any connection, stay on URL paste plus manual uploads; the script quality does not depend on a full install.
#Connecting Your Store
3.What permissions does the connection use?
Read access to products, variants, images, collections and the public-facing copy we need to draft scripts. It does not need permission to change prices, issue refunds, or read customer PII, and you should not grant a broader scope than the connect screen asks for. If a store’s permission model is locked down by an agency or a franchise HQ, you may only be able to import via URL. We do not use the connection to email your customers. If a procurement team wants a DPA before connecting, that is a reasonable enterprise request — do not connect a production store as a test on a personal laptop.
#Connecting Your Store
4.Can I connect more than one store to a workspace?
Yes, which is how agencies and multi-brand operators usually work. Keep stores labelled, and do not mix brand kits across them. One workspace per client is still cleaner when legal entities differ. Disconnecting a store stops new imports; videos you already rendered stay yours. It does not uninstall anything from the merchant’s Shopify admin beyond the connection you added. If two stores share SKUs, import carefully or you will generate ads against the wrong currency or the wrong product URL. Check the destination link on every creative before it goes to ads manager.
#Connecting Your Store
02
What Gets Imported
5.What product fields actually get imported?
Title, description, bullet-style body copy where we can parse it, price as shown on the page, compare-at price if present, gallery images, variant names, SKU identifiers, collection membership, and review text when it is on the page. Those fields feed the script generator and the default B-roll pool. We do not import inventory counts as a live stock feed, customer records, discount codes, or Shopify Flow. On a well-built product page this is usually six to twelve images plus enough copy to draft a 30-second ad without typing the benefits by hand.
#What Gets Imported
6.Are product images good enough to make the ad?
Often for a first test, rarely as the forever-asset. Lifestyle images beat ghost-mannequin packshots in UGC. If the gallery is one packshot on white, upload your own clips — especially for anything that has to move. Repeating a single hero image across 30 seconds looks like a placeholder and hold rate suffers. Delete seasonal or discontinued shots from the asset panel after import. Vertical lifestyle beats heavily cropped squares. You can still export 9:16, 1:1 and 16:9 at 1080p from mixed sources; the crop will be kinder if you start with taller frames.
#What Gets Imported
7.Does import include metafields, ingredients or size charts?
Only when they render on the public product page in a way the importer can read. Metafields that exist only in admin, app-powered size widgets, and tabbed content loaded by JavaScript are commonly missed. If a claim lives in a metafield and never appears in the description, do not expect the script to know it. Paste those facts into the product brief, or they will not make the voiceover. This is a regular surprise for stores that keep “real” copy in metafields and keep the description empty for theme reasons.
#What Gets Imported#Limits
8.Is the imported price always right?
It is the price the storefront showed at import time, for that market, in that currency, which may not be what a buyer in another country sees. Markets, B2B price lists, and automatic discounts will not all come through. Never treat the imported number as a live price API. If you speak a price in the ad, you own keeping it true. Many brands are safer saying “on the product page” than burning a number into a 30-second video they will reuse for six weeks. Re-import before a promo if the creative will mention a figure.
#What Gets Imported
03
Variants & Collections
9.How are variants handled — colour, size, scent?
Variants import as options under the product, with their images when Shopify has them. Generate against the variant you will land the click on, not against a random colourway. An ad for the sage hoodie that lands on a sold-out neon variant is a conversion leak. Size variants rarely need their own video unless fit is the story; colourways often do, because the B-roll has to match. You can batch colourway versions from the same script. Check swatch images — some themes use tiny sprites that look unusable as B-roll; swap those for lifestyle shots.
#Variants & Collections
10.Can I generate ads from a whole collection?
Yes. Pick a collection, select the SKUs you actually want in the batch, and generate rather than auto-making every object in “All products”. Unfiltered collections include gift cards, discontinued SKUs, and wholesale-only items you should not advertise. Use collections as a planning tool: “summer dresses” is a batch; 4,000 SKUs is a crawl. Scripts will be thin if collection products share a generic description. Spend ten minutes on the first product’s brief, then clone the structure. Catalogue scale without a brief is how you get 80 near-duplicates.
#Variants & Collections#Catalogue Scale
11.Do collection descriptions feed the script?
They can, but product-level copy should win. A collection blurb is usually marketing frosting — “shop the drop” — and makes a weak voiceover. Prefer the product description, review language, and a short brief you type. If all SKUs in a collection share one sentence of copy, write a proper brief once and reuse it. The importer will not invent fabric weight or a fit note that nobody wrote. Garbage in, UGC out. Collection copy is a fallback, not a strategy.
#Variants & Collections
12.What about combined listings, bundles and apps that rewrite the PDP?
Bundles, subscription widgets, and custom PDP apps often produce a page the importer only half understands — one hero image, a generic title, missing variant photos. Open the public URL in a private window and look at what actually loads. If the product is assembled by an app at runtime, upload assets manually and type the offer. Headless storefronts and Hydrogen custom storefronts are in the same bucket: the admin catalogue may be rich while the URL you paste is a JavaScript shell. Manual beats fighting a custom front-end.
#Variants & Collections#Import Failures
04
Catalogue Scale
13.Can I generate ads for an entire catalogue at once?
You can queue a large batch; you should not blindly render every SKU. Pick winners, newness, and SKUs with lifestyle images. A 200-product blast will consume credits, fill the queue, and produce near-duplicate talking heads. Start with 10–20 SKUs, 1–2 creatives each, then expand. Credits are charged on render, by length and resolution, not by import. Fifty free credits will not cover a full-catalogue render, and they are not meant to. Catalogue scale is a paid-plan workflow with a human choosing the list.
#Catalogue Scale
14.Will each product get a unique script?
It will get a script from its own title, bullets and reviews, which is unique only if those fields are unique. Stores that copy-paste one description across a colourway will get copy-paste ads. That is not a bug. Edit the brief for hero SKUs, and let long-tail SKUs share a structure with different B-roll. If you need 100 genuinely different angles, you need 100 different hooks, not 100 imports. See /knowledge-base/en/faq/bulk-generation-and-batch-workflows-faq for how to plan that without wasting the queue.
#Catalogue Scale
15.How do I keep 50 product ads on-brand?
Lock a brand kit, a shortlist of 4–6 avatars, a template, and a naming scheme before you batch. Collection-scale generation without those rails is how a store ends up with 50 unrelated presenters. Reuse on-body or in-use clips from a shared library. Export 9:16 first at 1080p; add 1:1 later for the SKUs that win. Review the first five renders fully, then let the rest run. If the first five are wrong, cancel the queue. Credits on a bad 50-product run are the expensive way to learn you needed a brief.
#Catalogue Scale
16.Can I sync new Shopify products automatically into ads?
Not as a fully automatic “new SKU, live ad, no human” pipeline, and that is deliberate. A new product still needs a hook choice, a claim check, and a look at the images. You can re-import the catalogue, see new SKUs, and generate. Treating Shopify as a firehose into paid social is how discontinued, out-of-stock and half-copied products get advertised. Use collections or tags to mark “ready to advertise” in your own process, then import that set. Automation without QA is a limit, not a missing checkbox.
#Catalogue Scale#Limits
05
Refreshing Data
17.How do I refresh product data after a price or copy change?
Re-import the product or refresh the connected catalogue, then check the asset panel before you re-render. Old images and old prices stick to projects until you refresh; they do not hot-reload into videos you already exported. If a promo price is in the voiceover, you must re-render — captions and audio will not change themselves. Build a habit: refresh before every launch batch, and again when packaging changes. Stale packaging in an ad that hits a new PDP is a conversion bug that looks like a media problem.
#Refreshing Data
18.Is there a live two-way sync with Shopify?
No. We read catalogue data; we do not write products, inventory, orders or discount codes back to Shopify, and we do not update ads when a SKU sells out. Inventory truth stays in Shopify. If you go out of stock, pause the ad in the ads manager. A two-way sync that auto-pauses paid social is not something we claim. This is one of the real limits of the integration: it is a creative import, not a merchandising system. Plan the operational pause yourself.
#Refreshing Data#Limits
19.How often should a store with frequent merchandising refresh?
Weekly if prices, bundles and hero images move; immediately after a packaging change; daily during a sitewide sale only for the SKUs whose ads speak a price. Refreshing everything every morning is wasted motion if your catalogue is stable. Assign one person to check the asset panel on the products that are actually spending. Agencies should put “PDP match check” on the QA list next to hook rate. The cheapest optimisation in Shopify-plus-ads is often deleting last season’s image, not rewriting the hook.
#Refreshing Data
20.Do already-rendered videos update when I refresh the store?
No. A rendered file is a file. Refreshing the catalogue updates what the next generation can see. If you need the new pack shot in a live ad, re-render and replace the upload in Meta or TikTok. This surprises people who treat the integration like a dynamic product ad feed. We export video files — 9:16, 1:1, 16:9 at 1080p, 4K on premium — not a live catalogue feed. Budget credits for those replacements, especially around promo windows. Unused paid credits roll over for one billing cycle; a sale week is what that rollover is for.
#Refreshing Data
06
Import Failures
21.Why did my Shopify import return no images?
Common causes: a password page, a geo-gated storefront, a theme that lazy-loads the gallery in a way we cannot see, app-based image hosts, or a product with only admin images not published. Open the URL logged-out, in the same market you are advertising. If you see images and we did not, upload them manually and keep moving — fighting a custom theme for an hour is worse than a three-minute upload. Fewer than three images should be treated as a failed scrape even if the title came through.
#Import Failures
22.The wrong product or language came through. What happened?
Markets, translated storefronts, and hreflang URLs. You pasted a .com page and got US copy while you wanted the /es listing, or a market redirect sent the importer to another catalogue. Paste the exact URL you want the buyer to land on, after any redirect. Disconnecting and reconnecting will not fix a wrong URL. Check title, currency and image against the PDP before you generate a batch. One wrong import multiplied by 20 SKUs is an expensive QA problem. Slow down at the URL step.
#Import Failures
23.Can I import from a Shopify store that is not published yet?
If the storefront is password-gated, the public importer will fail until you bypass that or connect with credentials that can read the catalogue. A coming-soon store with no product pages has nothing to scrape. You can still type a brief and upload sample photography. Do not wait on a public URL if you already have assets; the first test batch does not require a live checkout. Launch-week ads fail more often from missing B-roll than from a missing connection.
#Import Failures
24.HEIC images and weird file types from Shopify — will they upload?
The storefront images Shopify serves are usually JPG, PNG or WebP and those import. HEIC files from an iPhone dropped straight into the theme are a common rejection when you upload extras yourself — convert to JPG first. Very large PNGs and progressive formats can time out; compress and retry. The importer is not a DAM. If your product media lives in a third-party DAM and the PDP shows placeholders, we will import placeholders. Fix the PDP, or upload the real media in the project.
#Import Failures
07
Limits
25.What can the Shopify integration not do?
It cannot push ads live, cannot pause spend when you sell out, cannot rewrite your theme, cannot guarantee claim-safe copy, cannot import customer photos from reviews without you checking rights, and cannot turn a ghost-mannequin catalogue into fit-accurate fashion footage. It will not replace a product feed for dynamic ads. It will not bill Shopify instead of your Klip Kanvas plan — check /pricing for credits. Treat it as a fast, imperfect catalogue ingest. The human still picks the SKU, the hook and the truth of the claim.
#Limits
26.Does connecting Shopify include commercial rights to review photos?
No. A review photo on the PDP is not automatically cleared for paid ads. Customer faces, children, and third-party trademarks in those photos are your problem to licence. The importer may pull them as images; that is not a rights grant. When in doubt, exclude review photos from the B-roll pool and use brand assets. We supply the tooling; the advertiser owns compliance for everything that appears in the cut. This is the same rule as any other upload, and connecting the store does not weaken it.
#Limits
27.Is Shopify the only store I can import from?
No. Shopify pages scrape most reliably, then typical WooCommerce and BigCommerce templates, then many Amazon galleries. Custom single-page sites, login walls and geo-gates fail more often. You can always paste text and upload media. The connection experience is deepest for Shopify because that is where most of our catalogue-scale users live, not because other carts are forbidden. If you are on a headless setup, plan for manual assets. The first ad still only needs a product name, a few benefits, and one honest picture.
#Limits#Connecting Your Store
Ready to put this into practice?
Create your first AI UGC video ad in minutes — no filming, no actors, no editing.
