How much does manually re-typing Allegro orders into Subiekt cost?

It's past six in the evening. One desk in the office still has its lights on. Two monitors: on the left, the order list from Allegro (Poland's largest marketplace), on the right, Subiekt nexo (a Polish ERP by InsERT). The person in charge has a few dozen orders to re-type today, because a delivery came in this morning and nobody had time. Click the order, copy the buyer's details, create the customer, add the line items, issue the invoice, go back to Allegro, type in the tracking number. And again. And again.

I've seen this in several companies we've talked to. Nobody counts that hour, because it never shows up on any invoice. It shows up in overtime, in stock-level errors, and in the owner being afraid to let that person take a holiday. So I decided to do the math. Below is a calculation you can swap for your own numbers in fifteen minutes.

The math: what one order costs

Let's start with a single order. I wrote down everything that has to happen for an Allegro order to become a document in Subiekt and get back to the buyer with a tracking number:

• open the order in Allegro and check that it has been paid,
• find or create the customer in Subiekt: invoicing details, delivery address,
• add the line items and make sure each one is the same SKU that sits in the warehouse,
• issue the document: first a ZK (sales order), then an invoice or a receipt,
• go back to Allegro, change the status and type in the tracking number.

For a skilled person on a good day, that's 4 minutes. With a new customer, an order with three line items and a buyer's question in between, it's 6 minutes or more. For the math I'm using 5 minutes. That's an assumption, not a measurement from your company. If it's different for you, swap it.

The second number is the cost of an hour of work. Not net pay, but the full cost: gross salary, employer contributions, the workstation, holiday, sick leave. For an order-processing employee I'm assuming PLN 50 per hour. A reasonable range is PLN 45-60, depending on region and seniority. This is another assumption you can swap.

The third number is volume. I took the three levels we hear most often in consultations: 150, 500 and 1,500 orders a month.

The middle scenario is a typical store that sells on Allegro seriously but doesn't have a separate order-handling team yet. 500 orders a month, 5 minutes each. Over a year it comes to this:

per year, at 500 orders / month

500 h

per year, at PLN 50 / h (assumption)

PLN 25,000

of work, every single month

5 days

If you prefer a range: at 4 minutes and PLN 45 it's PLN 18,000 a year, at 6 minutes and PLN 60 it's PLN 36,000. All for the same 500 orders. 500 hours is roughly a quarter of one full-time employee's year, spent copying data from one window to another.

This calculation doesn't yet include a single error, a single return or a single invoice issued late. This is the cost of the re-typing alone, when everything goes well.

The costs that aren't in the table

The table only shows time. The real money disappears elsewhere. Below are the five places we see most often when we sit down with an owner to go over the process.

Stock levels. An order re-typed late is a product that still shows as "available" on Allegro even though it's gone from the warehouse. The result: selling something you can't ship, a cancellation, a bad review. One wrong SKU does the same thing in reverse: the system shows nothing in stock while the goods sit on the shelf.

Late invoices. A buyer asks for an invoice, the order person has a three-day backlog, the invoice goes out with a date that doesn't match the sale date. The accountant asks. Someone corrects it. Every such correction is more minutes that aren't in the calculation above.

Returns and credit notes. A manually issued invoice needs a manual credit note. With an Allegro return you also have to reverse the warehouse document and check whether the money went back to the buyer. It's the most thankless part of the job, so it usually sits the longest.

VAT OSS on EU sales. If you sell to buyers in other EU countries, once you cross the EU-wide threshold you account for VAT at the rate of the buyer's country. Manually, that means: remember the country, remember the rate, put the right one on every invoice. A mistake only surfaces at the VAT return, which is too late.

Knowledge in one head. The most expensive cost and the hardest to count. One person knows how customers are named, how numbering works, what to do with an order from a company without a VAT number. They go on holiday and for two weeks the orders wait, or the owner does them at night. Sick leave works the same way, just without warning.

Three ways: manual, subscription, your own code

From here, three roads lead out. Each makes sense in a different situation, and I won't pretend ours is always the best one.

Manual. Works as long as volume is low and the order person has other tasks that fill their day. At 150 orders a month it's a dozen or so hours. You can live with that, provided errors don't happen. The problem is that they do.

A subscription tool. BaseLinker (now Base.com) and similar integrators connect Allegro to Subiekt using a ready-made pattern: they pull orders, create documents, send statuses back. You pay monthly, configure it yourself or with an implementation partner, and it's up and running quickly.

Your own integration. Code written for your process, running on your infrastructure. You pay once for the implementation, the code is yours once it's paid for, and you extend it in the direction your business is heading, not the direction the tool's vendor anticipated.

When a subscription stops being enough

We usually see this a year or two after starting with an off-the-shelf tool. The company has grown, the process has become more complex, and the integrator still does what it did at the beginning. The symptoms are fairly repeatable:

• Multiple Allegro accounts: several brands, several legal entities, accounts for different countries, and each one needs to land in a different document series or a different warehouse.
• Non-standard numbering: separate series for the marketplace and the online store, the order number inside the invoice number, accounting requirements the template doesn't anticipate.
• Your own warehouse with locations, picking and batches, where an order has to reach a specific person in a specific zone, not just "go into Subiekt".
• Margin calculated before the listing goes live, after commission and shipping cost, using purchase prices from Subiekt. Integrators look at the order, not at profitability.
• Exceptions handled manually alongside the integrator: a company without a VAT number, a merged order, a prepayment from another channel. If there are several such exceptions a day, the integrator has stopped being an automation.

On top of that there's the bill for the subscription itself. I won't quote tool prices, because they change with every price list and depend on the plan. Do it yourself: take your monthly integrator invoice and multiply by 60. That's what five years costs. For example, if the subscription were PLN 300 a month, that's PLN 18,000; if PLN 600, that's PLN 36,000. These are placeholder numbers, not any company's price list.

Your own integration is a one-off cost for code that belongs to you once it's paid for. Add technical support if you want it, but there's no fee for "platform access". For a simple process, the subscription wins. For a process with exceptions and several accounts, the five-year math usually looks different. Run both before you decide.

What a custom integration looks like in practice

No jargon, step by step. This is what a typical Allegro-Subiekt nexo integration we build looks like.

Getting into Subiekt. Subiekt nexo has no public API in the sense Allegro does. What it has is Sfera: InsERT's official developer toolkit, available in the PRO edition. Through Sfera, the integration does in Subiekt exactly what a person would do: creates the customer, issues the document, checks stock. Just without the clicking. We cover where Sfera can surprise you in a separate post on the pitfalls of integrating with Subiekt nexo.

Pulling orders. The integration asks Allegro for new, paid orders every few minutes. It takes the buyer's details, the line items, the shipping cost, and whether the buyer wants an invoice. It matches line items to SKUs in Subiekt by your rules: EAN, manufacturer code, or a mapping we agree on at the start.

The document. The order becomes a ZK, i.e. a sales order, and from that a sales invoice or a receipt, depending on what the buyer chose. The numbering is yours: a separate series for Allegro, the order number in the notes field, whatever accounting requires.

Status updates. Once the document is ready and the parcel has been shipped, the integration sends the tracking number and status back to Allegro. The buyer sees the parcel is on its way, and nobody copies numbers between windows.

Error handling. Allegro doesn't respond, Subiekt is busy, the item isn't in the product catalog. The integration doesn't hide it. It puts the order in a queue, retries, and if it still can't get through, it says so specifically: this order, this problem, this person needs to act. There's no silent list of "didn't go through".

Idempotency. A hard word for a simple thing: the same order never enters Subiekt twice. Not after a restart, not after a power failure, not when Allegro returns the same order in two responses. Every order has its own identifier, and the integration remembers what it has already done with it. Without this, an automation makes more mess than a human.

What we've done for others

Three examples from our implementations, without client names.

For an Amazon seller operating in five EU markets we built an SP-API integration with Subiekt nexo PRO. Hundreds of invoices a month, each country with a different VAT rate, previously all done by hand. Today an invoice is created for every order and goes straight to accounting on its own. The implementation took 3 weeks. Details: Amazon EU, automatic VAT invoices.

At a home-appliance wholesaler with four price lists and 1,293 products, sales reps wrote down orders, someone re-typed them into Subiekt and manually checked which price list applied to the customer. A tablet app, a B2B portal and Subiekt nexo now work on the same data, and the company counts around 180 fewer hours of work a month. Write-up: B2B wholesaler.

A workwear store selling on Allegro listed offers one at a time and only calculated profitability in a spreadsheet after the sale. We built an application that publishes listings directly through the Allegro REST API and calculates margin after commission and shipping before publication; multi-account support is in code review. Zero manual listings. More: Allegro store.

How to start: run your own numbers

Before you talk to anyone, do the math yourself. It takes fifteen minutes and you'll get a number you can actually have a conversation about.

1. 01Measure the time per order. Don't ask, sit next to the person and time ten orders with a stopwatch. Take the average.
2. 02Check the order count for the last three months in the Allegro panel. Take the monthly average, not the best month.
3. 03Work out the fully loaded hourly cost of the person doing it: gross salary plus employer contributions, divided by hours actually worked.
4. 04Multiply: minutes times orders, divide by 60, times the hourly rate, times 12. That's your annual cost of the re-typing alone.
5. 05Add what happened last year, even roughly: how many stock errors, invoice corrections, days of backlog because of holidays.
6. 06Count how many exceptions you handle manually alongside the tool. If it's more than a few a day, the ready-made pattern no longer fits your process.

With that number in hand, I invite you to a free consultation. It's a 30-minute conversation: we walk through the process, check whether Subiekt and Allegro can be connected the way you want, and tell you whether it's worth it at all. If an off-the-shelf integrator is enough, you'll hear it straight.

If it is worth it, the next step is an audit for PLN 699 net. That's a few days of our work and an implementation document: integration architecture, exact scope, risks, a fixed price and a deadline. The full audit fee is deducted from the implementation price. A typical implementation takes 2-3 weeks from kick-off to production; a larger project is split into 2-3-week stages, and you get the specific timeline in the audit.