Ask a contact-centre team what is wrong with their platform and the answer is rarely the platform. Calls get answered, queues work, reporting exists. The frustration is with everything around it: the re-keying, the spreadsheets, the "we do it this way because the system can't".
That is the hidden cost of legacy integration, and it tends to be invisible on any budget line.
Where the cost actually sits
- Agent time spent copying information between systems after each interaction.
- Follow-up failures when the hand-off between systems depends on someone remembering.
- Workarounds that became process: macros, shared inboxes and manual reports that only work if everyone follows the steps.
- Change paralysis, because any workflow improvement touches several fragile pieces at once.
- Key-person risk, where the one integration that does exist was built by someone who has left.
None of these show up as "integration cost". They show up as headcount, error rates and projects that never start.
Why replacement is usually the wrong first move
The instinctive response is to replace the platform. Sometimes that is right. More often it is a large, risky programme that the business does not have the appetite for, and in the meantime nothing improves.
A replacement also does not fix the workflow problem on its own. If the surrounding processes are not redesigned, the new platform inherits the same workarounds within a year.
A more pragmatic route
Treat the legacy platform as a data source and a channel, and move the workflow out of it. Practically, that means:
- Wrap the old interfaces in a thin layer of integration services with modern, documented APIs. Expose only the operations the workflows need.
- Automate the hand-offs against those APIs, so that a completed call creates the right record and the right follow-up without re-keying.
- Retire the workarounds one at a time, each after its replacement has run reliably.
- Monitor the legacy interface, so a failure surfaces immediately rather than as a missing record days later.
The integration layer becomes an asset. When the platform is eventually replaced, the services are re-pointed and the workflows carry on.
What good looks like
The test is simple: can a workflow change be made without touching the legacy platform? If the answer is yes, the platform has stopped being the constraint, and the replacement decision can be made on its merits rather than under pressure.
Where cloud fits
Azure, or an equivalent, is a natural home for this kind of integration layer: small services, managed messaging, monitoring and identity are all available without building infrastructure. The point is not the cloud itself but the ability to put a modern, owned layer between the workflow and the old system.
We have written up one such engagement as a case study.
- Contact centres
- Legacy systems
- Integration
- Azure