A ready-made e-commerce platform or B2B portal from an external vendor runs on a subscription model in which the fee isn't fixed: above a certain threshold, it's charged as a percentage of revenue — the value of sales the company processes through the platform. The storefront, customer accounts, and order history are stored on the vendor's infrastructure. The price list looks moderate as long as revenue stays low. The bill changes as the company grows: the more it sells, the more it pays for the platform, even though the platform does nothing more.
The analysis below is based on the current price lists of three popular platforms — Shopify Plus, BigCommerce, and Salesforce Commerce — as of June 2026, with a reference to the Polish platforms IdoSell and Shoper. It shows what a ready-made platform really costs in zloty and dollars, where the cost invisible in the price list surfaces, and what alternative a portal built as a company-owned system represents. The thesis of the whole piece is this: for a B2B company with growing revenue, tying itself long-term to a platform billed on revenue is a costly decision, because the fee is linked to the value of sales, not to the value the platform brings to the company.
A platform in this category is delivered as SaaS — the company doesn't buy the system, it rents access, paying a fee that, above a threshold, depends on revenue. The system runs in the vendor's cloud, and the storefront, catalog, B2B customer accounts, negotiated individual price lists, and order history are all saved on its infrastructure.
This model produces recurring traits that show up in the pricing of leading platforms. The fee is tied to revenue: for some, as a percentage of sales value above a threshold; for others, as an automatic move to a higher, more expensive tier once an annual sales cap is crossed. Payment and transaction commissions come on top of the platform fee, charged separately on every order. B2B features — advanced pricing, credit limits, multi-branch handling — are often paid add-ons. The storefront, customer data, and order history stay with the vendor. Each of these traits sounds neutral on its own. Combined and tied to revenue, they add up to a bill that grows along with the company's sales.
The real cost of a ready-made B2B platform for a company doesn't depend on a single number, but on the revenue the company processes through it and on the way the vendor chooses to link the fee to sales. The table below compares five platforms as of June 2026 — the figures are net prices and change over time.
| Platform |
Origin / currency |
Billing model |
Starting point |
Link to revenue |
| Shopify Plus |
US / USD |
subscription + percentage above threshold |
~USD 2,300–2,500/mo (~PLN 8,400–9,100) |
0.25–0.40% of GMV above threshold, capped at USD 40,000/mo; +0.20% for external gateways |
| BigCommerce |
US / USD |
tiers based on annual sales |
~USD 1,500–3,000/mo for mid-market |
automatic move to a higher tier once annual revenue is exceeded |
| Salesforce Commerce |
US / USD |
percentage of GMV + license |
from ~USD 60,000/yr (~PLN 220,000) |
~1–3% of GMV plus license; implementation USD 200,000–500,000 |
| IdoSell |
Poland / PLN |
subscription + "cost of success" |
subscription from ~PLN 249/mo |
surcharges for orders above the package + revenue commission ("pay as you grow" model) |
| Shoper |
Poland / PLN |
flat subscription, no commission |
~PLN 299/mo (regular price) |
no percentage of sales; cost independent of revenue |
Net prices, as of June 2026. Foreign platforms bill in dollars; converted at an exchange rate of approximately PLN 3.68 per dollar, which means the zloty bill also depends on the exchange rate. Salesforce's percentage rates and Shopify's thresholds are sometimes negotiated and not always disclosed before signing the contract.
Shopify Plus. The subscription starts at USD 2,300 a month on a three-year contract, or USD 2,500 on an annual one, covering revenue up to around a million dollars a month. Above that threshold, the platform fee switches to a percentage model: 0.35% of revenue on a three-year contract or 0.40% on an annual one, charged as whichever is higher than the subscription, capped at USD 40,000 a month. Using an external payment gateway adds 0.20% on every transaction. On top of that come card commissions and app costs, which means the real bill is often several times the subscription alone.
BigCommerce. The platform bills in tiers based on annual sales, at roughly USD 1,500–3,000 a month for mid-sized companies. A defining feature of the model is the automatic move to a higher, more expensive tier once the annual revenue cap is exceeded — a company can't stay on a cheaper plan once its sales grow. BigCommerce doesn't charge its own transaction commission when using an external payment gateway, but the platform's cost still rises with revenue achieved.
Salesforce Commerce. This platform applies the purest revenue-linked model: the fee is a percentage of sales value, typically 1–3% of GMV, plus a license. The cost for a mid-sized company starts at around USD 60,000 a year and grows with sales — for example, a store with USD 25 million in annual revenue at a 2% rate pays around USD 500,000 in commission plus the license. On top of that comes an implementation in the USD 200,000–500,000 range and annual price increases of five to nine percent.
IdoSell. This Polish platform (IAI S.A.), which serves both B2B and B2C sales, bills in zloty, using a subscription plus what it calls the cost of success — surcharges for orders above the package and a revenue commission. The company itself describes it as a "pay as you grow" model: the fee grows as the business grows. Billing in zloty removes the exchange-rate risk present with foreign vendors, but the logic of tying the fee to revenue stays the same.
Shoper. The second Polish platform bills a flat subscription and explicitly charges no sales commission — the cost is independent of revenue, regardless of whether the company sells tens or hundreds of thousands of zloty a month. Fees for apps, integrations, and online payments remain, but the link between platform cost and sales volume disappears. Shoper shows that a revenue tax is a choice of model, not an unavoidable feature.
The scale of the cost only becomes visible once the fee is set against revenue. In a percentage model, doubling sales doubles the platform portion of the bill, even though the platform delivers nothing more in return. The cost doesn't scale with the value the system brings to the company, but with the value the company generates on its own.
| Annual revenue (GMV) |
Salesforce Commerce (~2% + license) |
Shopify Plus (flat / 0.35%) |
| PLN 5M |
~PLN 320,000 (~USD 88,000) |
~PLN 110,000 (~USD 30,000) |
| PLN 15M |
~PLN 520,000 (~USD 142,000) |
~PLN 110,000 (~USD 30,000) |
| PLN 30M |
~PLN 820,000 (~USD 225,000) |
~PLN 110,000 (~USD 30,000) |
| PLN 60M |
~PLN 1,420,000 (~USD 389,000) |
~PLN 210,000 (~USD 58,000) |
Chart caption: the higher the revenue, the higher the bill — even though the platform does nothing more. A flat-subscription model (the Polish Shoper without commission, or an in-house system) would appear here as a flat, horizontal line, independent of revenue. Model rates; billed in dollars, converted at an exchange rate of approximately PLN 3.68 per dollar, as of June 2026.
What the company pays for: a commission on its own success
A revenue-linked model has a trait that sets this category apart from the rest: the company pays more precisely when it's doing better. Sales growth — the result of the sales team's work, marketing, and product quality — pushes up the platform bill, even though the platform contributed nothing more to that growth than what it was already delivering. The fee becomes a form of tax on the company's own success.
This mechanism also works the other way, and it can sting hardest at sales peaks. A record-revenue month — during a busy ordering season, say — means the highest platform bill, right when the company needs its margin the most. On top of the percentage of revenue come payment commissions charged on every transaction and, with automatic tier escalation, the inability to stay on a cheaper plan once the sales cap is crossed. The bill therefore grows on several fronts at once, each tied to how much the company sells.
A Polish angle: the same model, sometimes without the percentage
Polish platforms show two different answers to the same model. IdoSell, a Polish platform serving B2B sales, uses a subscription plus a cost of success — surcharges for orders above the package and a revenue commission, described outright as a "pay as you grow" model. The logic here is identical to that of foreign vendors: the fee grows with sales. Being Polish only removes the exchange-rate risk, since billing happens in zloty.
Shoper, the second Polish platform, takes a different path: it charges a flat subscription and takes no percentage of sales. The platform's cost is independent of revenue here. This shows that tying the fee to sales is a choice of the vendor's business model, not a necessary condition for running a store. But that only solves one problem. A flat subscription without commission still means renting: the storefront, customer accounts, and order history stay on the vendor's infrastructure, and the company remains subject to its decisions on pricing, limits, and the platform's future. The percentage of revenue disappearing doesn't remove the rental or the dependency.
The cost the price list doesn't show
The most serious cost of a ready-made platform never appears on an invoice. It's the dependency on a single platform and the loss of control over the storefront, customer data, and the sales relationship. It's this layer that turns a pricing decision into a strategic one.
The storefront, catalog, individual B2B price lists, customer accounts, and order history are all built within a structure imposed by the vendor and stored on its infrastructure. The longer a platform is used, the more sales relationships and purchase data accumulate within it, and the harder switching becomes. Migrating to another platform is a separate, costly project — the total cost of switching an enterprise-class platform is counted in the hundreds of thousands, and in the millions of zloty for the most extensive deployments. The vendor's decisions become the company's problem: a product's sunset forces migration on a fixed deadline, as shown by the end of support for on-premise SAP Commerce deployments (version 2205) in July 2026, which puts companies still on that version in front of a required migration. On top of that come contractual price increases in the range of five to nine percent a year.
The result is a situation in which the sales channel and the relationship with B2B customers — a resource that determines the company's survival — sits in the infrastructure of an outside party, on terms set by the vendor and dependent on its decisions.
Data as a strategic resource in the age of AI
Data on B2B orders, buyers' purchasing patterns, and price history has stopped being just material for sales reports. Today it's fuel for a company's own AI tools: a product recommendation engine tailored to a specific customer's purchase history, demand forecasting at the SKU and buyer level, automatic personalization of price offers. Each of these applications works more precisely the longer and richer the transaction history in the company's system is. That history can't be recreated or bought — it's a resource built up over years of the company's own sales, permanently unavailable to competitors.
The decision about where this data lives stops being a technical matter. Entrusting it to an external vendor's infrastructure means a resource of growing strategic importance stays outside the company's control — accessible on the vendor's terms, not portable without a costly migration, and exposed to changes in the platform's policy. A company planning to bring AI into its own sales process should factor this dependency into the equation at the platform-selection stage, not after years of use.
What changed the equation: a lower barrier to building software
The case for building an in-house B2B portal would have sounded unreasonable just two years ago. Building a sales platform from scratch meant a long development project, a high cost for the development team, and a timeline measured in months. For most B2B companies, renting a ready-made platform was then the only sensible answer, and the percentage of revenue was the price for avoiding that risk.
AI tools supporting software development have changed this equation. Coding assistants based on large language models cut application build time many times over, shifting much of the work from manual coding to design, validation, and integration. An area that once required a multi-person team and a multi-month schedule is now handled by a smaller team in a fraction of the time. The barrier to building a dedicated B2B portal — with its own catalog, price lists, and integration with the company's ERP — has dropped enough that, for a growing number of businesses, it has stopped being an obstacle.
The change has a second layer. The same AI tools that speed up building the system also become a built-in feature of it. A dedicated B2B portal is built faster thanks to AI used as a development tool, and at the same time it contains AI running during operation — for example in product recommendations, order forecasting, and handling buyer inquiries. As a result, the cost-and-time argument that spoke against building an in-house platform for years has largely stopped holding up.
The alternative: a B2B portal as a system the company owns
The alternative to a revenue-billed subscription is a dedicated B2B portal built as a system the company owns, matched to its sales process and running on its own infrastructure. The catalog, individual price lists, credit limits, and order handling mirror the real model of working with buyers, and the portal integrates with the company's ERP and other systems. The storefront, customer accounts, and order history stay with the company. A built-in AI layer handles recommendations, forecasting, and inquiry handling.
The difference is structural, not cosmetic. In the subscription model, the company bears a cost tied to revenue, growing with sales and independent of whether the platform brings any new value in return. In the dedicated model, the company bears the cost of building a portal that then belongs to it — with no percentage of revenue and no bill that grows with sales.
Fairness requires stating this plainly: a dedicated portal isn't free and doesn't remove every fee. Payment handling still involves a payment processor's commission, charged on transactions regardless of the chosen solution, and the portal itself requires the cost of building, hosting, and maintenance. The difference lies in what the company is paying for. Instead of the percentage of revenue a platform charges for mere access, the company funds the creation of its own system, whose cost doesn't grow with sales and whose storefront, data, and customer relationships stay under its control. That such a system can be built is shown by working web applications created this way — dedicated tools with their own back end, maintained as company systems rather than extensions of ready-made platforms.
A ready-made B2B platform is a rational choice under certain conditions, and this article doesn't argue otherwise. It works well when revenue is small and stable, when the sales process is standard and fits within ready-made constraints, when a fast start without a build phase matters, and when the company doesn't treat where customer data is stored or the revenue-linked cost as a strategic issue. For such an organization, a subscription — especially a Polish one on a flat model without commission — is often the simplest and sufficient answer.
The equation flips once at least one of the following conditions is met. Revenue is growing, so the sales-linked model starts working against the company and the percentage of revenue exceeds the cost of maintaining an in-house system. The B2B sales process is unusual — complex pricing, credit limits, branch handling — and fitting it in requires constantly working around a ready-made platform's constraints. Customer and order data is sensitive material, and where it's stored is subject to legal requirements or a security policy. The portal is meant to be integrated with the company's ERP and other systems as part of a coherent ecosystem. The decision horizon spans several years, and the cost summed over that time exceeds the cost of building an in-house solution. In these situations, a dedicated portal stops being the more expensive solution and becomes the cheaper and safer one over the long run.
A subscription B2B platform — regardless of whether the vendor is Polish or foreign — tempts with a single fee, but as the company grows it reveals three costs: a fee tied to revenue and growing with sales, payment commissions and add-ons charged separately, and dependency on a platform that locks away the storefront, customer data, and the sales relationship. Some vendors drop the percentage of revenue, like the Polish Shoper, but the rental and the dependency remain. The most serious cost concerns data: at a time of AI's growing role, the relationship with B2B customers and order history remain a valuable company resource, and the subscription model keeps them outside the company's control.
A dedicated B2B portal built as a company-owned system reverses this logic. The cost of building it and a platform-independent payment processor's commission replace the percentage of revenue, the storefront and customer data stay on the company's own infrastructure, and the portal mirrors its real sales process instead of forcing it into someone else's framework. A full description of this approach and the company's other dedicated web applications is available at artechconsult.com/solutions. The starting point for a conversation about an in-house portal is a free consultation.
The prices cited in this article come from vendors' official price lists and independent pricing analyses as of June 2026, are given in net prices, and change over time. Foreign platforms bill in dollars; figures are given in dollars with a zloty equivalent converted at an exchange rate of approximately PLN 3.68 per dollar, and the zloty bill additionally depends on the current exchange rate. Percentage rates (especially for Salesforce Commerce) and revenue thresholds are sometimes negotiated and not disclosed before signing the contract, so the amounts given are a model estimate, and the actual cost depends on the contract and the company's revenue.
