1C and ERP Integration with a Web App: What to Put in the Spec Before Signing

The company ordered a “convenient employee portal.” Six months later, accounting still enters the same shipments into 1C manually because “integration was moved to phase 2.” Development budget is spent, adoption is low, and IT explains that “1C doesn’t expose an API.”

Integration with 1C, SAP, or another ERP is not a technical footnote on the last page of the spec. It is a decision about where the truth lives for items, counterparties, and order statuses. Below is a framework for CEOs and CTOs: what to lock in before signing the estimate so you do not end up with two parallel accounting systems.

Why “we’ll add it later” almost always costs more

Typical chain:

• a shipment request is created in the portal;
• the warehouse confirms in Excel or chat;
• accounting re-enters data into 1C “the usual way”;
• discrepancies show up at month-end close.

One extra hour per operator on 50 shipments a day is ~200 hours per month. At fully loaded cost of $15–25/hour, that is $3,000 – $5,000/month in hidden loss — before errors and penalties to partners.

Integration cannot be “added for free” after the portal architecture and data model are fixed. Define the exchange contour in Discovery — at least as a list of entities and master/slave rules.

Three questions the CEO should ask before signing

1. Who is master for each entity? Items, counterparties, stock, order statuses — one source of truth per entity.
2. How often must data match? Real-time, every 15 minutes, nightly batch — technology and budget depend on this.
3. What happens on mismatch? Alert, block the operation, manual reconciliation — the rule belongs in the runbook, not “on a call.”

Master data: who is in charge

While the answer is “both in 1C and in Ivan’s Excel” — any integration will be contentious. One directory owner and an update rule are mandatory before development (see also the digitalization checklist for CEOs).

How to connect 1C and a web app technically

For most projects with 50–500 users, one of these approaches (or a combination) is enough:

• HTTP services / REST in 1C:Trade, ERP — publish methods like “create order”, “get stock”.
• OData — convenient for reading directories; writes need separate methods with validation.
• Message queue (RabbitMQ, Kafka) — when you need async delivery guarantees between ERP and microservices.
• File exchange (scheduled XML/JSON) — legacy but still common; plan monitoring and idempotency.

For the CTO, the point is not a “fashionable stack” but an exchange contract: API version, error format, retry, dead letter queue, logging of every sync with a correlation id.

Real-time vs batch: when to choose what

A common mistake is demanding “everything in real-time” for all entities. That multiplies integration cost 2–3× without proportional benefit.

Minimum spec for integration

Integration budget ranges (2026)

Figures after Discovery; highly dependent on 1C version, configuration customizations, and entity count.

• Directory read + one-way document export — from $2,000 – $4,500, 3–6 weeks.
• Two-way exchange (requests, statuses, counterparties) — from $5,500 – $12,000, 1.5–3 months.
• ERP + WMS + portal + queues — from $13,000 – $32,000, 4+ months.

Comparison: 200 hours of manual re-entry per month × 12 ≈ $36,000 – $60,000/year in operational loss — integration often pays back in the first year without “magic ROI.”

Bottom line

1C/ERP integration is not “one more API.” It is an agreement on a single source of truth, sync frequency, and behavior on mismatch. Lock this in the spec before signing — otherwise “phase 2” becomes a separate project with rework of the portal you already built.

Map your contour — automation and integrations, corporate portal, or request a consultation.

Related: 5 signs Excel is no longer enough, 7 intranet modules, digitalization: CEO checklist, IT contract: 10 CEO questions.