Custom Marketplace Integration vs Connector: How to Choose
Connector or custom marketplace integration? An honest guide for multichannel sellers: when each fits, how pricing works, hidden costs, and a checklist.
- Written by
- Md AshikSenior Software Engineer, Artisan Dojo
- Published
- Reading time
- 8 min read
Quick summary30-second read
- Choose a connector if your ERP and marketplaces are supported and your processes are standard; go custom when the connector cannot model your business.
- Custom work pays off for unusual ERPs, complex rules like bundles or per-channel stock, unsupported marketplaces, or fees growing faster than your margin.
- Connector costs usually scale with order volume or GMV, while custom costs scale with complexity and change, so compare total cost over three years.
- Count the hidden costs: manual workarounds and lock-in for connectors, and API changes, monitoring and maintenance for custom builds.
- Many sellers run a hybrid, using a connector for standard flows and custom code for the gaps, with one source of truth per data type.
If your ERP is a common one, your marketplaces are mainstream and your processes are fairly standard, an off-the-shelf connector or multichannel platform is usually the faster and cheaper choice. A custom integration starts to make sense when the connector cannot model how your business works, when your ERP or a marketplace is not supported, or when connector fees grow faster than your margin. Many sellers end up with a hybrid: a connector for the standard flows, plus a small custom service for everything it does not cover.
We build custom integrations for a living, so we have a bias. This guide tries to be honest anyway: custom work where a connector would do is just expensive software you did not need.
What is the difference between a connector and a custom integration?
The terms get used loosely, so here is what we mean:
- Multichannel platform or connector: a product that sits between your systems and the marketplaces and handles listings, stock, prices and orders through ready-made connections. Examples include ChannelEngine, Tradebyte, Linnworks and Rithum (formerly CommerceHub, which acquired ChannelAdvisor). You configure it rather than program it.
- ERP with built-in marketplace connections: some e-commerce ERPs ship marketplace connections as part of the product. PlentyONE (formerly plentymarkets) is a typical example. Many other ERPs offer marketplace plugins from the vendor or third parties.
- Custom integration: code written for your setup that talks directly to the marketplace APIs, such as Amazon SP-API, Kaufland's Seller API or Zalando's partner APIs, and to your ERP or warehouse system. You own the logic, and you are responsible for keeping it running.
- Hybrid: a connector or plugin handles the standard flows, and custom code fills the gaps.
When is a connector the right choice?
A connector is often the better answer, and we tell clients so. It usually wins when:
- Your systems are supported out of the box. If the connector has maintained connections to your ERP and every marketplace you sell on, most of the hard work is done.
- Your processes are standard. Orders in, stock and prices out, tracking back. If that covers almost everything, you are buying a solved problem.
- You want to test new marketplaces quickly. Platforms with large channel networks let you try a new marketplace or country without a development project for each one.
- You have no technical team. The vendor handles API changes, hosting and monitoring of its own connections. That is real value if nobody in-house can do it.
- Your volume is still modest. Usage-based pricing is gentle when you are small, and building custom software before you know a channel works is premature.
When does a custom integration make more sense?
Custom work pays off when the connector becomes the constraint rather than the shortcut. Common signs:
- Your ERP is unusual. Legacy, industry-specific or heavily customized ERPs often have no ready connector, or only a generic file import that loses information.
- Your business logic does not fit. Bundles and kits, complex variant structures, per-channel stock allocation, B2B and B2C on the same items, split shipments from several warehouses, or rules like "never list below landed cost plus fees" are hard to express in configuration screens.
- A marketplace you need is not supported, or the connector supports it only partly, for example orders but not returns, or listings but not the attributes your category requires.
- Volume and speed matter. At high volumes you may need control over sync frequency, rate limit use and how fast stock changes reach each channel.
- Fees grow with you. When pricing is tied to GMV or order volume, cost rises as you grow, even though the integration's work barely changes.
- You want to own your data and logic. Mapping rules, order history and error logs live in your own systems, so switching tools later does not mean rebuilding your processes.
Top tip
Before deciding, list the five things your current setup handles badly or not at all. If a connector demo cannot show each of them working with your own data, assume you will need custom work for them anyway.
What drives the cost of each option?
Prices depend on volume and scope, and most vendors here quote individually, so instead of numbers, here is how each side charges.
Connectors and multichannel platforms usually combine some of these:
- Onboarding or setup fee, often depending on how complex your integration is.
- A recurring license or subscription. The basis varies by vendor. ChannelEngine, for example, says its license fee depends on marketplace GMV, SKU count and scope, plus a success-based fee on GMV growth, and that it does not charge per marketplace. Linnworks, by contrast, says its plans are tiered by monthly order volume, with no percentage fees.
- Add-on modules for features like warehouse management, forecasting or advanced listing tools.
- Overages when you exceed the volume your plan includes.
Custom integrations cost money in a different shape:
- Build: a one-time project to design, develop and test the integration. Scope drives it: how many marketplaces, how many data flows, and how messy your ERP's interface is.
- Hosting: servers, queues and databases. Usually a small line item.
- Maintenance: ongoing work to keep up with API changes, fix edge cases and add small features. This is the cost people forget.
The core trade-off: connector costs tend to scale with your business volume, while custom costs scale with complexity and change. A seller with high volume and stable processes often finds custom cheaper over a few years. A seller with modest volume, many channels and frequent changes often finds the opposite.
What are the hidden costs?
Both options have costs that do not show up on the first quote.
On the connector side:
- Workarounds. When the connector cannot do something, people fill the gap by hand with spreadsheets and manual edits. That staff time is a real cost.
- ERP-side connection. Many platforms still need a connection to your ERP. If yours is not supported, that piece is custom work anyway.
- Lock-in. Your mappings, rules and listing data live in the vendor's system. Moving later means rebuilding them.
On the custom side:
- API changes. Marketplaces version, deprecate and change their APIs on their own schedule. Amazon, for example, has been retiring older listing feed formats in favor of JSON-based APIs. Someone has to read those notices and update the code in time.
- Monitoring. You need logging, alerting and someone who reacts. Otherwise the first sign of trouble is a customer complaint.
- Knowledge risk. If one freelancer built it and left, nobody may understand it.
This is why integration maintenance belongs in the decision, not after it.
Connector vs custom vs hybrid: comparison table
| Factor | Connector / multichannel platform | Custom integration | Hybrid |
|---|---|---|---|
| Time to first orders | Fast if your systems are supported | Weeks to months, depending on scope | Fast for standard flows, custom parts follow |
| Upfront cost | Low to moderate onboarding | Highest: a build project | Moderate |
| Ongoing cost driver | GMV, order volume or plan tier | Maintenance and hosting | Both, but smaller of each |
| Fit for unusual ERPs and logic | Limited to what is configurable | Full control | Custom code covers the gaps |
| New marketplace coverage | Broad, if the channel is in the network | One integration per marketplace | Connector for breadth, custom for gaps |
| API change handling | Vendor's job | Your job or your developer's | Split |
| Data and logic ownership | Mostly in the vendor's system | Fully yours | Mixed |
| Best for | Standard processes, many channels, small team | High volume, complex logic, unsupported systems | Mostly standard needs with a few hard requirements |
When is a hybrid approach the best answer?
In practice, hybrids are common. Typical patterns we see:
- A connector handles listings and orders, and a custom service pushes stock from a warehouse system the connector does not support.
- The ERP's marketplace plugin covers Amazon and eBay, and a direct integration handles a marketplace it does not cover.
- A connector runs day to day, and custom code pulls settlement and fee reports into the finance system or a data warehouse.
You keep the connector's breadth and pay for custom work only where it earns its keep. The risk is two systems both thinking they own the same data. Decide the source of truth for each data type before you start.
Decision checklist: connector or custom?
The more of the first group applies to you, the more a connector fits. The more of the second, the more custom or hybrid fits.
Points toward a connector:
- A maintained connector exists for my ERP or shop system.
- Every marketplace I sell on, or plan to, is supported with the flows I need.
- My order, stock and pricing processes are fairly standard.
- I have no developer or integration partner to own custom code.
- I want to test new channels quickly, without a project per channel.
Points toward custom or hybrid:
- My ERP is legacy, industry-specific or heavily customized.
- I rely on business rules that connector demos could not reproduce.
- I need a marketplace or data flow the connector does not support.
- Usage-based fees are growing faster than the value I get from them.
- I want my integration logic and data in systems I control.
- Staff spend real time every week on manual workarounds.
FAQ
Is a custom marketplace integration more expensive than a connector?
Upfront, almost always. Over several years it depends on volume and complexity. Compare total cost over three years, including staff time spent on workarounds.
Can I start with a connector and switch to custom later?
Yes, and it is a sensible path. Keep your own records of product mappings and business rules, so a later move does not start from zero. Many sellers replace one flow at a time.
Who maintains a custom integration when marketplace APIs change?
Whoever you agree on before go-live: an in-house developer or an external partner on a maintenance plan. It should never be "nobody".
Do I still need custom work if my ERP has a marketplace plugin?
Often not. If the plugin covers your marketplaces and flows, use it. Custom work becomes useful when the plugin misses a marketplace, a data flow such as returns or settlements, or a business rule you depend on.
Not sure which way to go?
We help sellers make this call before they commit money, and we will say "use the connector" when that is the right answer. If custom or hybrid fits better, we build marketplace ERP integrations and keep them running afterwards. You can see all our services, or describe your ERP, marketplaces and the problems you are running into below, and we will tell you what we would do.