TERM · SOFTWARE & INTEGRATIONS

System integration

What is system integration?

System integration is the work of connecting two or more programs — a website, warehouse software, accounting, a CRM — so that data moves between them automatically instead of being retyped. The integration is the link and its rules: which system owns each field, what is sent, and what happens when a message fails. Synchronisation is that same link running again and again, on a schedule or on an event.

Example

A shop with 4,000 products takes about 40 orders a day. Each one is typed into the warehouse software by hand, and stock levels are uploaded from a file in the evening. At three minutes an order, that is two hours a day of typing, and between the evening upload and the next morning the shop keeps selling quantities it no longer has.

Once the systems are integrated, an order reaches the warehouse the moment it is paid, the warehouse sends back a dispatch number and a waybill, and stock changes travel back to the website every few minutes. The typing disappears. What is left is the exceptions: an order the warehouse refused, a product whose code has changed.

Integration and synchronisation are not the same thing

Integration is the project: which systems connect, through which interface, which one owns the price and which one owns the quantity, how products are matched, and what happens when something fails. Synchronisation is the running of it — the same exchange repeating on a schedule, say every 15 minutes, or on an event. The event version usually uses a webhook: the system where something happened sends the message itself instead of waiting to be asked. Most links go over an API, some over files, and the most fragile ones reach straight into another system’s database, which breaks at its next update.

Why it matters for a business

Integration decides how many people a process needs and how often a customer is told something untrue. It also has a life after launch: other systems ship new versions, keys expire, formats move. So the question is not only whether the link starts working, but who watches it and how long it takes to notice that it has stopped.

We build these links as part of the web software we develop, most often between a shop, a warehouse system and accounting.

What to ask

  • Which system owns each field — price, quantity, description, order status?
  • What are products matched on, and what happens when there is no match?
  • How often does the exchange run, and is there an event-driven path for the urgent parts?
  • What happens to a failed message — is it retried, where is it recorded, who is alerted?
  • Is there a log of what was sent and accepted, and how far back does it go?
  • Who maintains the link when the other system changes version?

A term you do not recognise? Write to us and we will add it.

The glossary grows with the questions we are asked.