Every marketplace onboarding deck promises vendors a new sales channel. Then it asks them to export a spreadsheet, re-map their categories to yours, keep two price lists in sync by hand, and re-enter stock every morning. That is not a sales channel, it is a second job, and good vendors decline it. The onboarding model that actually scales does the opposite: the vendor keeps their existing store as the single source of truth, and the marketplace reads from it. This is how we onboard vendors onto a live multi-vendor marketplace without any of them changing their own workflow.
The never-re-enter rule
The one principle that governs everything downstream: a vendor should never have to type a product into your system that already exists in theirs. Their WooCommerce or Magento store, or the product feed they already generate for shopping ads, is authoritative. Your job is to ingest it, enrich it, and present it, not to become a second place they maintain the same data.
This is not just kindness. Every manual re-entry step is a place where the two systems drift out of sync: a price they changed on their shop but not on yours, a product they discontinued that still shows as in stock on the marketplace. Drift is a trust problem, and on a marketplace trust is the whole product. Reading directly from the vendor's live store means their truth is your truth, on every sync.
Connect the store they already have
In practice, vendors arrive in one of three shapes, and each has a clean seam to read from:
- WooCommerce. The Store API exposes products, variants, prices and stock without a bespoke integration. It has sharp edges — most notably VAT-inclusive pricing and shipping rates that only appear at the cart — which we cover in adding WooCommerce stores to a marketplace.
- Magento. The public GraphQL endpoint is the seam, and it comes with its own quirks — attribute sets, numeric option IDs — walked through in ingesting a Magento catalog.
- A product feed. If a vendor already produces a feed for shopping ads, that feed is a ready-made ingestion source. The discipline that keeps feed ingestion reliable — scheduling, idempotency, change detection — is its own topic in product feed ingestion that doesn't fall over.
The point of supporting all three is that the vendor picks nothing. You look at what they already run and read from it. None of the vendors on our marketplace changed how they operate their own store to join.
Six independent vendors — supplement brands, a physiotherapy store, an equipment retailer — run on our marketplace, and not one of them altered their internal workflow to be there. They keep selling on their own shop exactly as before; the marketplace is an extra shelf their catalog appears on, kept current by ingestion, not by their staff.
Sync scheduling: nightly full, hourly partial
Reading a store once is onboarding; reading it on a schedule is operations. Two cadences cover almost everything:
| Sync type | Cadence | What it catches | Cost |
|---|---|---|---|
| Full | Nightly | New products, deletions, restructured categories, anything the partial missed | Higher — reads the whole catalog |
| Partial | Hourly | Price and stock changes on products already known | Low — touches only what moved |
The nightly full sync is the safety net: it reconciles the whole catalog so nothing gets permanently stuck. The hourly partial keeps the fast-moving fields — price and availability — fresh through the day without the expense of a full reread. The trick that makes the hourly cadence cheap is change detection: a price update should not re-run expensive enrichment, because it did not change anything a shopper reads. More on that mechanism in the ingestion piece.
Mirror shipping policy, don't invent it
A multi-vendor cart charges shipping per vendor, and the fastest way to make a vendor distrust the marketplace is to charge their customers a shipping fee that does not match their own shop. So we do not invent shipping numbers. We mirror each vendor's real fees and free-shipping thresholds, verified against their live store rather than assumed from a form they filled in.
Getting this right is fiddly — one vendor ships cross-border with weight-based rates that have to be crawled from its own checkout because there is no flat number to copy — but the principle is simple: the marketplace's shipping math should be a faithful mirror of policy the vendor already stands behind. We go deep on the mechanics in per-vendor shipping.
Pre-launch seed checks
Before a vendor's catalog goes visible to shoppers, it passes a short list of checks. These are the ones that catch the launch-day embarrassments:
- Category coverage. Every source category maps to your shopper-first taxonomy, or the products land in a sensible default — never in an empty or mislabelled bucket. Unmapped categories show up as empty facet chips, which look broken.
- Images present. Products with no image are gated out at ingest rather than shipped as grey placeholders. A catalog full of missing images reads as abandoned.
- Stock signal sane. Confirm the store actually reports availability, and that "no signal" defaults to a safe value rather than silently marking everything out of stock (or worse, in stock).
- Price and currency. Confirm whether the source prices are VAT-inclusive or exclusive before a single product is listed, because getting it wrong is invisible until a customer is overcharged.
- Shipping mirrored and verified. The vendor's fees and thresholds match their own shop, checked, not guessed.
Run these as a gate, not a suggestion. A vendor's first impression of the marketplace is their catalog on launch day; a garbled one is hard to walk back.
Why vendors say yes
Vendor acquisition is the hard half of any marketplace, and the ingest-don't-ask model is the strongest argument you have. A vendor evaluating your marketplace is really asking one question: what does this cost me in ongoing effort? If the honest answer is "connect your existing store once, keep running it exactly as you do now, and receive orders", the incremental-channel maths becomes easy for them to accept. There is no new inventory system, no duplicate price list, no daily stock chore. That is also why audience-first marketplaces solve supply so much more cheaply than platforms that start empty — the demand is already there, and the ask of the vendor is near-zero. We unpack that dynamic in solving the cold-start problem.
Key takeaways
- Ingest, don't ask. Read the vendor's existing WooCommerce, Magento or feed store; never make them re-enter data they already maintain.
- Two cadences cover it: a nightly full sync as the reconciling safety net, an hourly partial for fast-moving price and stock.
- Mirror shipping policy, verified against the vendor's own shop — a fee that does not match theirs destroys trust faster than any bug.
- Gate go-live on seed checks: category coverage, images present, sane stock, correct VAT handling, mirrored shipping.
- The near-zero effort ask is your best pitch — vendors accept an incremental channel that changes nothing about how they already work.
Frequently asked questions
How do you onboard a vendor to a marketplace without spreadsheets?
How often should a marketplace sync a vendor's catalog?
What checks should you run before a vendor goes live?
Why would a vendor agree to join a marketplace?
Onboarding vendors shouldn't feel like a migration.
We add a managed multi-vendor marketplace to a site that already has traffic, and onboard vendors by ingesting the stores they already run.
Request early access See the live marketplace →