Repeat orders
The customer reorders the same items or a predictable variation. If every order is a project, the site is there to request a quote, not to sell.
Blog · E-commerce
B2B e-commerce: when it pays off (and when it doesn't)
Selling online to business customers is not B2C with a login page. It pays off when orders repeat and the cost of handling them by hand is higher than the cost of digitising them. Here is how to tell before you invest.
B2B is not B2C with a login page
In B2C the customer arrives, chooses and pays. In B2B the person often cannot buy: they have a budget, a negotiated price list and an internal approval to respect.
A B2B portal has to live with rules that B2C does not know: different prices per customer, tiered discounts, minimum order quantities, payment terms, multiple delivery addresses, credit limits, approvals and a salesperson following the order. If the site ignores those rules, it produces orders that the back office has to correct by hand. At that point you have added a channel, not removed work.
That is why the project is not measured in pages. It is measured in orders that arrive without human intervention, errors avoided and hours given back to the people who currently retype data from an email into an ERP.
The useful question is not “how much does a B2B e-commerce site cost”. It is “how many orders do we handle, how long does each one take, and what happens if the customer orders alone”. The answers change the project more than any platform choice.
Six conditions that make B2B e-commerce pay off
You do not need all of them, but the more are true, the more the project makes sense.
A stable, coded catalogue
Codes, units of measure, packs and availability are clear. Without clean data, an online catalogue amplifies errors instead of reducing them.
Customers who know what they want
The order does not require advice. If advice is the value, the site can prepare the request, not replace it.
A measurable manual cost
Email, phone, PDFs and re-keying take hours. The saving is measured in those hours, not in website visits.
A reason to serve outside office hours
Customers order when the sales team is away. A portal extends the service window without adding office staff.
Growth beyond the territory
New customers far away can buy without a trip. The site becomes the first point of contact, not a replacement for the relationship.
When it does not pay off, and what to do instead
The wrong portal costs more than no portal: it adds a channel and leaves the manual work untouched.
- Few customers and few orders. Build and maintenance costs never come back. A good PDF catalogue and a request form are better.
- Every order is configured to measure. You need a quote configurator or a request workflow, not a cart with standard prices.
- The customer wants to talk to a person. The site can collect information and prepare the conversation, not replace the relationship.
- The catalogue has no owner. First decide who updates codes, prices and availability, then digitise.
- Logistics and payments are not ready. The portal takes orders the back office cannot fulfil, and the customer notices immediately.
- Low margins and high returns. Handling costs can exceed the benefit, especially if every order needs a human check.
B2B e-commerce is not an IT project. It is a process change. If the process is not defined, software only makes it more visible.
Four models, not one
“Build a B2B e-commerce site” can mean very different things. Choose the model based on who buys and how many rules they bring.
| Model | What it does | Operational complexity | When it makes sense |
|---|---|---|---|
| Catalogue with quote request | Shows products, hides or shows prices, collects a structured request. | Low | The price depends on the customer, the quantity or the configuration. |
| Private order portal | Registered customers see price lists, stock and history; they order on their own. | Medium | Repeat orders and negotiated price lists. |
| Hybrid | Standard items are bought online; configurations and non-standard quantities go through a quote. | Medium-high | A mixed catalogue, part standard and part custom. |
| Open catalogue or marketplace | Many customers, onboarding, payments and standardised logistics. | High | You want to reach customers you do not know and cannot handle every order by hand. |
Are you ready? The checklist before the quote
A missing item is not always a blocker: it is work to do before or during the project.
- Customer recordsEvery customer has a code, a price list, payment terms and delivery addresses.
- Price lists and discountsYou know who sees which price and where the rule comes from. The rule should live in one place only.
- AvailabilityThe warehouse is current and accessible, or the site will show promises you cannot keep.
- Units and packsDo you sell boxes, pallets, metres or pieces? The catalogue has to say so without ambiguity.
- Orders and reordersCan the customer repeat a previous order or upload a list of codes?
- Payments and creditBank transfer, card, direct debit, credit limit: the rules must be explicit before the first order.
- Shipping and returnsWho pays for what, with which carrier, in what time and through which return process.
- SupportWho answers when an order does not go through? The portal does not remove support, it moves it.
The calculation nobody does before choosing a platform
Before comparing suppliers, measure the current cost. You do not need a complicated model.
Monthly order-handling hours = orders per month × minutes per order / 60. If an order takes twelve minutes across email, data entry and checking, and you handle three hundred a month, that is sixty hours a month. At that point the question is not “how much does the portal cost”, but “how much does continuing like this cost”.
Then add the cost of errors: wrong lines, re-sent shipments, correction time, phone calls. The saving is not only speed: it is less work to redo. And add the customer’s time: an order sent at six in the evening does not wait until the next day.
On the other side, put the total cost: platform, integration, maintenance, training and internal time to run it. A portal nobody uses is a cost, not a saving. The decision comes from comparing those two numbers, not the monthly subscription.
Replace the numbers with yours. The example shows the method. If you do not know the minutes per order, measure a sample week: it is the most useful data you can bring to a first conversation.
Integration is the project (the front end is the easy part)
The difference between a useful portal and double data entry is the connection to the ERP, the warehouse and accounting.
Without integration, the order arrives in the portal and someone retypes it into the ERP. You have digitised the collection, not the process. The customer sees a modern site; the back office keeps doing the same work, with one extra step.
You need a data owner: someone who decides the item code, the price list, the availability. If the same data lives in three places, the portal will show three different truths. Choose the source first, then connect everything else.
API or file? An API is preferable when the ERP exposes one and volumes are meaningful. A periodic file can be enough at low volume, but it needs checks and a person responsible for it. Either way, plan for errors: partial orders, negative stock, blocked customers, failed payments. Testing happens on the edge cases, not the happy path.
A simple volume calculation
A simple sum to understand whether a portal makes sense.
| Item | Value |
|---|---|
| Orders per month | 120 |
| Minutes per order | 8 |
| Hours saved per month | 16 |
Formula. Orders per month × minutes per order / 60 = hours saved.
Frequently asked questions
Is B2B e-commerce only for large companies?
No. The deciding factor is not size, but how repeatable the orders are and what it costs to handle them by hand. A small company with many reorders can benefit; a large one with a few complex orders may not.
Can I use Shopify for B2B?
Shopify has B2B features, but the real question is whether your model fits its rules: price lists, payments, credit, ERP integration. If you need a lot of custom behaviour, adaptation costs can outweigh the benefit.
Do I need a customer login area?
Not always. If prices are public and orders are standard, an open catalogue may be enough. A login area is needed when price lists, availability or terms depend on the customer.
How do I handle different prices for each customer?
With price lists and customer groups fed by the ERP. The rule should live in one place: if the price lives in three systems, they will diverge sooner or later.
How long does it take?
It depends on the catalogue, integrations and the number of customers to import. A first portal with catalogue and orders can start in a few weeks; testing with customers takes longer than building.
How do I measure whether it works?
Orders moved to the portal, back-office hours saved, data-entry errors, fulfilment time and the share of active customers. Visits are the last metric, not the first.
Start from the orders, not the platform
Tell us how many orders you handle, who enters them and what happens today when a customer wants to buy. We will work out whether a portal pays off and where to start.