Simple Site, Online Store or a Platform Integrated with 1C? The Differences Nobody Explains Before You Pay

In short. There are five levels of website, and the difference between them isn't the design — it's how much manual work they remove. A landing page brings you an enquiry. An online store brings you orders that you then retype into 1C by hand. An integrated platform asks nothing of you: the order lands in 1C by itself, stock decrements automatically, the customer receives an invoice, and you see everything in your CRM. The price difference between level 3 and level 4 is typically three to four times. The difference in time saved is 40–60 hours a month. This guide shows you where you are, when it's worth moving up, and what breaks along the way.

Contents

1. Why "how much does a website cost" is the wrong question
2. The five levels — full comparison
3. Signs you've outgrown your current level
4. What "integrated website" actually means
5. 1C integration: how it really works
6. One direction or two? The difference that changes everything
7. Who owns the truth: the most important technical decision
8. CRM integration
9. Payment integration
10. Ten real integration pitfalls
11. The right architecture for an integrated platform
12. How a project runs: stages and timelines
13. Decision tree: which level you need
14. Frequently asked questions

1. Why "how much does a website cost" is the wrong question

It's the first question we get and, unfortunately, the least useful.

A website can cost 150 euros or 15,000. Both figures are correct, because "website" doesn't describe a product — it describes a category as broad as "vehicle". A bicycle and a lorry are both vehicles.

The question that produces a useful answer is different: how many hours a month does your team lose doing work the website could do by itself?

A real example, typical of a Chișinău store with 800 products and 20 orders a day:

Fifty hours a month is a half-time employee doing nothing but copying data from one system into another. At average retail wages, that's a recurring cost which exceeds the price of integration in under a year.

And that's before counting the errors: the order typed in wrong, the product sold when it was already out of stock, the old price left live for three days.

2. The five levels — full comparison

Level 1 — Landing page

What it is: one page, one message, one button.

What it does: presents an offer and collects a contact. That's all.

Who it's for: an ad campaign, a product launch, a single service, an event.

What it doesn't do: no catalogue, no user accounts, no complex payments, connected to nothing.

When you outgrow it: when you have more than one service to present, or when customers start asking things that should already be on the site.

Level 2 — Brochure site

What it is: 5–15 pages — services, about, portfolio, blog, contact.

What it does: makes you findable in Google, builds credibility, generates enquiries through a form.

Who it's for: service firms, practices, agencies, manufacturers selling through distributors, restaurants.

What it doesn't do: doesn't sell directly, doesn't track customers, doesn't know your stock.

When you outgrow it: when you're taking more orders by phone and Instagram than fit in a notebook.

This is where most companies in Moldova sit. And for many of them it's entirely sufficient — not every business needs an online store.

Level 3 — Standard online store

What it is: product catalogue, cart, checkout, online payment, admin panel.

What it does: sells around the clock without you touching each order.

Who it's for: retail, physical or digital products, a catalogue of up to a few thousand lines.

What it doesn't do: it doesn't talk to your accounting system. The order appears in the store's admin panel. From there, somebody copies it into 1C by hand. The stock figure on the site is whatever you typed in. So are the prices.

When you outgrow it: when order volume turns manual copying into a full-time job. The threshold usually hits somewhere between 15 and 30 orders a day, depending on product complexity.

Level 4 — Integrated store

What it is: the level 3 store, plus automatic connections to the systems you already run.

What it does: - products, prices and stock arrive from 1C automatically - orders leave for 1C automatically - payment confirms automatically and marks the order - the customer lands in the CRM automatically - the invoice is issued automatically - order status updates in both systems

Who it's for: stores with 500+ products, a real warehouse, accounting in 1C, and a team that should be selling rather than copying data.

What it doesn't do: it doesn't run complex manufacturing and doesn't replace a full ERP.

When you outgrow it: when you add multiple warehouses, multiple sales channels, or business logic no off-the-shelf system covers.

Level 5 — Custom platform

What it is: software built around your process rather than the reverse. It may include a B2B customer portal, a product configurator, per-customer pricing, approval workflows, supplier integrations.

Who it's for: distributors, marketplaces, businesses whose model doesn't fit any ready-made platform.

The comparison table

Remember the manual-work row. That's the entire difference between level 3 and level 4 — and the real reason it's worth paying three times as much.

→ Level 2 as a service: Online store — complete eCommerce platforms with payments and delivery built in.

3. Signs you've outgrown your current level

You don't need an audit. Here are the signs, in ascending severity:

Sign 1. Someone on your team has "enter website orders into 1C" among their duties. That isn't a task — it's a symptom.

Sign 2. You've sold a product that was already out of stock. More than once. You called the customer to apologise.

Sign 3. The price on the site differs from the price in 1C, and nobody is quite sure which one is right.

Sign 4. When someone's on holiday, orders pile up unentered until they return.

Sign 5. You can't quickly answer "how much did we sell online last month, by category". The data sits in two places and doesn't reconcile.

Sign 6. You've given up updating part of the catalogue on the site because it's too much work.

Sign 7. A customer rings about their order and you have to open three systems to answer.

Three signs ticked means you're losing money every month. Five means the integration would already have paid for itself.

4. What "integrated website" actually means

Everyone uses the term loosely. Here's what it means concretely, following a single order through the system.

In a non-integrated store (level 3):

Customer places an order ↓ Order appears in the site's admin panel ↓ [ HUMAN ] opens 1C and retypes the order ↓ [ HUMAN ] checks the bank statement ↓ [ HUMAN ] marks the order paid on the site ↓ [ HUMAN ] issues the invoice in 1C ↓ [ HUMAN ] emails the invoice ↓ [ HUMAN ] decrements site stock so it isn't sold twice

Seven human interventions. Each takes minutes and each can be done wrong or forgotten.

In an integrated store (level 4):

Customer places an order ↓ Order goes to 1C automatically, as a document ↓ Payment confirms by webhook, order is marked paid ↓ 1C reserves stock and issues the invoice ↓ The invoice reaches the customer automatically ↓ Updated stock returns to the site at the next exchange ↓ Customer and order appear in the CRM

Zero human interventions for a normal order. People step in only for exceptions — returns, cancellations, a customer who changed their mind.

That's the difference. Not the design, not the speed, not the page count. How many human decisions a normal order requires.

5. 1C integration: how it really works

In Moldova, 1C is the de facto standard for accounting and inventory. Practically every company with a warehouse runs it, in one of the versions adapted to the local market.

The good news: the exchange with your website doesn't need inventing. There's a standard.

CommerceML — the shared language

CommerceML is an XML format developed by 1C for exchanging commercial data. It's the language in which 1C and the website talk to each other.

The mechanism, simplified:

1. 1C generates XML files containing products, prices, stock and images
2. It sends them to an exchange endpoint on the site (a dedicated URL behind authentication)
3. The site processes the files and updates the catalogue
4. In the other direction, the site exposes new orders in the same format
5. 1C collects them and creates the corresponding documents

The exchange runs on a schedule — typically every 30 minutes to 2 hours, depending on volume and business needs.

What flows from 1C to the site

What flows from the site to 1C

6. One direction or two? The difference that changes everything

This is the decision that separates a two-week project from a two-month one. And it's the decision most clients make without understanding what they're choosing.

Practical advice: if you're under 10–15 orders a day, one-way synchronisation removes 80% of the pain at a third of the cost. Entering ten orders takes twenty minutes, not four hours.

Above that threshold, the second direction becomes mandatory.

The real-time myth

Clients often ask for "real-time synchronisation". In practice that's rarely what they actually want and almost never what they need.

A full catalogue exchange for a few thousand products takes minutes and loads both systems heavily. Run every ten seconds, it would take down both the website and the 1C server.

The correct answer is hybrid:

• catalogue and prices — scheduled, every 1–2 hours
• stock — more often, every 15–30 minutes, for active products
• orders — sent immediately, on placement
• payment confirmation — immediately, by webhook

That way the critical things happen instantly and the bulky ones don't block the system.

7. Who owns the truth: the most important technical decision

If you take away one technical thing from this article, make it this one.

For every field that exists in both systems, you must decide from the outset which system is the source of truth. Without that decision the integration produces chaos — one system constantly overwrites the other's changes, and nobody understands why prices change by themselves overnight.

The typical split that works:

Note the rows for description and categories. This is the most expensive mistake in badly executed integrations: everything gets synced from 1C, and at the first exchange the site loses all its SEO-optimised copy, replaced by dry item-catalogue names.

Three months of content work, erased overnight. We've seen it more often than we'd like.

8. CRM integration

If 1C answers "what did we sell", the CRM answers "who am I talking to and what happens next".

What syncs in a proper integration:

• The contact is created automatically on the first order or first enquiry
• Order history is visible directly on the customer record
• Lead source carries through: which campaign, which page, which keyword
• Abandoned carts trigger a task or an automated sequence
• Segmentation is built on real data: lifetime value, frequency, categories bought

The benefit felt soonest is a different one: correct sales attribution. When you know a campaign produced 40 orders at an average of 900 MDL, your ad budget decision stops being a guess.

Without this link you have Google Analytics telling you how many "conversions" you had, but not how many turned into money received. Those are very different things.

9. Payment integration

In an integrated platform, payment isn't just a button — it's the event that triggers the entire processing flow.

The correct architecture has three layers:

1. Payment methods. Card, instant payment, cash on delivery, bank transfer. The more you offer, the fewer abandoned orders. The order in which you display them matters as much as having them.

2. Webhook confirmation. The processor notifies your server when payment succeeds. That is the only source of truth — not the customer's redirect back to the site, which anyone can forge by opening that URL manually.

3. Propagation through the system. Payment confirmation triggers, in sequence: marking the order, sending it to 1C, reserving stock, issuing the invoice, notifying the customer, updating the CRM.

For a detailed comparison of payment methods in Moldova and the fee savings arithmetic, see our guide to MIA Instant Payments for online stores.

→ Our dedicated payments guide: MIA Instant Payments for online stores — fees 3x lower than cards.

10. Ten real integration pitfalls

These are the problems that actually occur. Not theory — things that wreck projects.

1. The large catalogue that kills the exchange. Beyond 5,000 lines, XML processing hits PHP's time and memory limits. The answer is chunked processing that resumes where it stopped, not endlessly raising the limits.

2. Duplicates. On the second exchange, products get created again instead of updated, because matching is done on name rather than a stable identifier. Result: a doubled catalogue. Always match on the 1C code or GUID.

3. Loss of SEO content. Optimised descriptions on the site are overwritten with item-catalogue text. Prevented by the decision in section 7.

4. Incomplete status mapping. 1C has one set of document states, the site has another. If the mapping is incomplete, orders land in a state neither system understands and sit there.

5. Unreserved stock. The customer pays, but stock only decrements at the next exchange. In the gap, another customer buys the last unit. Reservation has to be immediate, on payment confirmation.

6. Heavy images. Images in 1C are typically unoptimised. Transferred as-is, they destroy your site speed and Core Web Vitals. They need automatic resizing and conversion on import.

7. Encoding and characters. Romanian diacritics and Cyrillic corrupt if encoding isn't handled properly. It's immediately visible and looks unprofessional.

8. Simultaneous editing. Someone changes a price in 1C, someone else on the site, within the same hour. Without the source-of-truth rule, the last exchange wins — arbitrarily.

9. No exchange log. When something fails to sync, without a log you have no chance of finding out why. The log isn't optional, it's part of the deliverable.

10. No test environment. The exchange is tested directly against the production database, and the first error corrupts real data. You need a separate environment with a copy of the database.

→ For levels 4–5 see our service: Web and mobile applications — custom SaaS platforms, CRMs and ERPs.

11. The right architecture for an integrated platform

Without jargon, here's what a build that holds up looks like:

The site stays a site. You don't load accounting logic into it. The site sells, presents and collects orders.

1C stays the system of record. That's where real stock, official prices and documents live.

Between them sits an exchange layer — not a direct connection. That layer translates formats, remembers what has been sent, retries what failed and keeps the log. If 1C is down for maintenance, orders aren't lost — they queue and go through when it returns.

Every operation is idempotent. Sending the same order twice doesn't create two documents. This is guaranteed by unique identifiers checked on every processing run.

Everything is logged. What was sent, when, with what response. Without that, debugging is guesswork.

There's alerting. If the exchange fails three times in a row, somebody gets notified. You don't discover the problem a week later when a customer asks about their order.

12. How a project runs: stages and timelines

For a level 4 integrated platform, a realistic project looks like this:

Total: 10–20 weeks.

Stage 1 is the one clients want to skip to get to "the real work" sooner. It is precisely the stage that decides whether the project succeeds. An integration started without field mapping gets rebuilt from scratch in month three.

13. Decision tree: which level you need

Answer in order:

Do you sell products directly from the site? → No, services only: level 1 or 2. Stop here, you don't need the rest. → Yes: continue.

How many orders a day do you have, or expect? → Under 10: level 3. Manual entry is bearable. → 10–30: continue. → Over 30: level 4, no debate.

Do you keep records in 1C or a similar system? → No, in Excel: level 3, but sort your records out first. You can't integrate a spreadsheet. → Yes: continue.

How many products in the catalogue? → Under 200 and rarely changing: level 3 is enough. → Over 200, or prices and stock changing often: level 4.

Do you have processes that fit no standard store? Per-customer pricing, approvals, a configurator, a B2B portal? → Yes: level 5.

14. Frequently asked questions

What does a website integrated with 1C mean? A website that exchanges data automatically with the 1C software: products, prices and stock come from 1C to the site, and orders from the site land in 1C as documents, without manual entry.

What is CommerceML? An XML format developed by 1C for exchanging commercial data between an accounting system and a website. It's the standard used by most 1C integrations with online stores.

Does synchronisation happen in real time? Usually not, and it doesn't need to. Catalogue and prices sync on a schedule every 1–2 hours, stock more frequently, and orders and payment confirmations go through immediately. A full real-time exchange would overload both systems.

Can 1C be integrated with WooCommerce? Yes. There are established solutions implementing the exchange over CommerceML, in one or both directions. The choice depends on your 1C version, catalogue size and process complexity.

How long does a 1C integration take? One-way synchronisation: 2–4 weeks. Two-way: 6–12 weeks. Add the time to build the website itself, if it doesn't already exist.

What happens if the 1C server is down? In a correct architecture, orders queue and are sent once 1C returns. Nothing is lost. In a poorly built integration, orders from that window disappear.

Will I lose my site's SEO copy after integration? No, provided it's agreed from the start that descriptions and commercial categories belong to the site, and only prices, stock and item-catalogue data come from 1C. It's one of the most common mistakes in rushed integrations.

Can I move from my current store to an integrated one without rebuilding it? Often yes, if the current platform is a mainstream one and well built. A technical audit determines whether integration is feasible on the existing structure or whether rebuilding is cheaper.

Is integration worth it at 5 orders a day? Usually not. Five orders is ten minutes of manual entry. Invest in sales until you reach a volume where automation pays for itself.

What matters more: design or integration? They depend on each other but answer different questions. Design decides whether the visitor buys. Integration decides whether you still have time to run the business after they have.

How we work

We build websites and integrated platforms for businesses in Chișinău and across Moldova — from brochure sites to stores connected with 1C, CRM and online payments.

What we do differently:

• We start with analysis, not design. The first deliverable isn't a mockup, it's the field map and the source-of-truth decision for each one. That document prevents most of the problems above.
• We build an exchange layer, not a direct link. With a queue, automatic retries, logging and alerting.
• We test against a copy of the live database, including failure scenarios, not just the happy path.
• We never overwrite your SEO content. Ever.
• We hand over the exchange documentation. If you work with someone else tomorrow, they'll understand what was built.

We work with WooCommerce, with custom Laravel platforms, and with 1C in the versions used in Moldova.

👉 Request a free technical review — we'll look at what you have, tell you which level you need and how long it takes. No obligation, and no attempt to sell you level 5 when level 3 is enough.

The timelines and thresholds in this article are indicative, based on typical projects. Your specifics depend on your 1C version, catalogue size and process complexity.

Last updated: 23 September 2026