Key takeaways
- Extraction is automated and takes seconds at any size. Almost all the cost is human time.
- Price should flatten sharply as catalogues grow. Pricing that scales linearly with product count is a warning sign.
- The redirect map is the most commonly omitted line item and the one that protects your traffic.
- Ask what happens in week two. That answer separates a service from a data transfer.
Ask three providers to migrate the same 5,000 product store and you can get $300, $1,200 and $9,000. That spread is not a market failing to find a price. It is three different jobs wearing the same name.
What actually costs money
Moving the data is the cheap part, and has been for years. A catalogue of any size can be read and reformatted in seconds. What consumes real time:
| Task | Scales with | Automatable? |
|---|---|---|
| Reading the source catalogue | Nothing meaningful | Fully |
| Writing the target format | Nothing meaningful | Fully |
| Field mapping | Number of distinct fields | Partly |
| Variant restructuring | How odd the source model is | Barely |
| Category and navigation rebuild | Tree depth, not product count | Barely |
| Redirect mapping | Number of indexed URLs | Partly |
| Verification | Risk tolerance | Partly |
Notice how little of that grows with product count. A 50,000 item catalogue does not need ten times the mapping of a 5,000 item one: it has the same fields, usually the same variant model and a similar category tree. It just has more rows flowing through the same pipe.
Use this as a test. Ask a provider what their price would be at ten times the size. If it is roughly ten times the money, they are either doing by hand something that should be automated, or charging for volume because volume is easy to put on an invoice.
The four ways to buy it
Do it yourself
Cash cost: nothing, or a plugin licence.
Reasonable for a few hundred products with a simple variant model. The risk is not the export, it is the parts that fail quietly: a truncated file that looks complete, images that stop resolving when the old host is cancelled, a variant model that collapses wrong. Our guides cover the traps for WooCommerce to Shopify and the reverse if you want to attempt it.
A self-service migration tool
Cash cost: typically tens to a few hundred dollars, priced per record.
The tool moves the data. You still decide how variants collapse, whether categories become collections or tags, what the redirect map looks like, and whether the result is correct. That is a fair trade when the catalogue is straightforward.
Watch the pricing model: per-record pricing is exactly the linear scaling described above, so these get expensive precisely where automation should have made them cheap.
A done-for-you migration service
Cash cost: usually a few hundred to a couple of thousand.
Someone else runs the extraction, does the mapping, builds the redirects and verifies the result before handing it over. You are buying the judgement calls and the checking, not the data transfer.
A full agency project
Cash cost: five figures and up.
Appropriate when the migration is genuinely the small part of a rebuild: new design, new integrations, ERP or POS syncing, multi-region. If someone quotes five figures to move a catalogue and nothing else, ask what the rest of the money is doing.
See your actual number in seconds
Paste your store URL. We count the catalogue and give you a fixed price on the spot, no email and no waiting.
The line items that get left out
When comparing quotes, these are where the difference usually hides:
- The 301 redirect map. The single most important deliverable for anyone with existing search traffic, and the one most often quietly excluded. Without it every product URL 404s on launch day.
- SEO fields. Meta titles and descriptions live outside the standard product export on most platforms. Losing them is invisible until rankings drift.
- Image re-hosting. Exports carry image URLs, not files. If nobody confirms the destination has pulled them in, your images die with the old hosting.
- Custom fields. Specifications, catalogue references, compliance data. Default exports drop them.
- Post-launch fixes. Ask directly: if something is wrong on day eight, is that included?
- Order and customer history. Frequently quoted, rarely useful. Consider archiving instead.
Five questions that price a quote honestly
- Is the redirect map included, and do you build it from Search Console or from a crawl? Search Console knows about pages a crawler cannot reach.
- What happens to products that break the target platform's limits? Shopify caps at 3 options and 100 variants. There is no correct automatic answer, so someone has to decide.
- How do you verify? "We check it" is not a method. Product counts, zero missing images, variant totals and a real test order are.
- What is the price at ten times the size? The shape of that answer tells you how much is automated.
- What is not included? The most useful question you can ask, and the fastest way to make two quotes comparable.
How we price it
We publish fixed prices by catalogue size, and the curve flattens hard because the underlying cost does. Extraction is automated, so a larger catalogue costs us little more than a small one; what we are actually charging for is the mapping, the redirects and the verification.
Every price includes products, variants, images, categories and SEO fields, verified before handover. You get the number on screen in seconds by pasting your store URL, rather than waiting a day for an email.
If your catalogue is genuinely unusual, very large, multi-language, or on a platform that publishes nothing, that is a real quote rather than a band, and we will say so instead of guessing.