Salesforce and Dynamics 365 integration patterns

Many organisations run Salesforce for sales and service and Microsoft Dynamics 365 for finance and operations. Each is strong on its own. Together, they only deliver value if customers, products, orders and invoices stay consistent across both.

Decide who owns which data

The most important integration decision is not technical. For every entity, agree which system is the source of truth and which systems only read it. A typical split looks like this:

• Accounts and contacts: created in Salesforce, credit and billing details mastered in Dynamics 365.
• Products and prices: mastered in Dynamics 365, published to Salesforce.
• Opportunities and quotes: Salesforce.
• Orders, invoices and payments: Dynamics 365, with status shared back to Salesforce.

Choose the right pattern for each flow

Not every flow needs to be real-time. Matching the pattern to the business need keeps the integration simpler and cheaper to run.

• Real-time APIs for actions users wait on, such as checking stock or credit during a quote.
• Event-driven messaging for changes that should propagate quickly, such as an order being confirmed.
• Scheduled batch for high-volume, low-urgency data such as historical invoices or price lists.

Pick an integration layer

Point-to-point connections work for one or two flows but become fragile as you add systems. Microsoft-centric landscapes often use Azure Integration Services — Logic Apps, Service Bus and API Management. Organisations with many systems or strong reuse goals often choose MuleSoft for API-led connectivity. Either way, put monitoring, retries and alerting in from day one.

Design for failure — Integrations fail — networks drop, records are locked, validation rules change. The difference between a good and a bad integration is whether failures are visible, retried safely and easy to fix.

Common pitfalls

• Duplicate customers because both systems can create accounts.
• Mismatched product codes, units or currencies.
• Integrations built around today's fields with no plan for change.
• No owner for monitoring and support after go-live.

Where to start

Map the end-to-end process — usually quote-to-cash — agree data ownership, then integrate the highest-value flow first. Prove it in production, then extend to the next flow.