Executive Summary
In short:
- The problem: Most Australian organisations run critical operations on software that is years old. When nobody is clearly accountable for patching, fixing and improving it, small faults turn into outages, security gaps widen and every change request feels risky.
- The numbers: The Australian Signals Directorate (ASD) received over 84,700 cybercrime reports in FY2024–25. The Office of the Australian Information Commissioner (OAIC) recorded 1,205 data breach notifications in 2025, a record high and up 8% on 2024. Malicious or criminal attacks caused 716 of them.
- The pattern in practice: The case studies in this article show two things that rarely appear in a project brochure: long-running support of a multi-system environment, and ongoing care of a connected healthcare platform where the systems on either side keep changing.
- The framework: Good application maintenance and support follows six steps: inventory, baseline, prioritise, support, improve and document. Five questions will tell you quickly whether a provider can do it.
What's next?
List the three applications your business cannot afford to lose, and write down who is accountable for each one, when it was last patched and who could take over tomorrow. That single page is the most useful input to any decision about support.
Why Does Your Most Important Software Always Seem to Break on a Monday?
It is 8.15 am on Monday. A payment run will not process, a report that finance relies on is blank, and the person who understands the system is on leave until Thursday. Someone finds a workaround in a spreadsheet. By lunchtime the workaround has become the process.
Nobody planned for this. The application was built years ago, it worked, and it quietly slipped out of anyone's job description. Patches were deferred because testing them felt like a project. Small improvements were postponed because the original developer had moved on. Each decision was reasonable. Together they created a system that everybody depends on and nobody owns.
That is the gap application maintenance & support services are meant to close: a clear, accountable arrangement for keeping live software secure, reliable and useful as the business changes around it.
This article uses case studies from C9, an Australian custom software developer, to show what that looks like on real projects. It also sets out the figures that explain why the risk is rising, a practical framework you can apply this week and the questions that separate a credible provider from a confident one.
In plain terms: Application maintenance and support is the ongoing work of keeping live business software healthy. It includes fixing defects, applying security patches, improving performance, supporting users and adding features as the business changes.
What Do Application Maintenance & Support Services Actually Include?
The phrase covers four distinct kinds of work. Software engineering standards, including ISO/IEC 14764, describe them as corrective, adaptive, perfective and preventive maintenance. Knowing the difference helps you ask for the right thing.
| Type |
Plain-English meaning |
Example |
| Corrective |
Fixing faults after they appear |
A report that shows the wrong total |
| Adaptive |
Keeping software working as its environment changes |
Updating an integration when a connected system changes |
| Perfective |
Improving performance or adding features |
Speeding up a slow query or adding a new screen |
| Preventive |
Reducing the chance of future faults |
Patching, refactoring and monitoring |
Support sits alongside these. It is the human side: helping users, triaging problems, escalating the serious ones and communicating clearly while a fix is under way.
Many organisations buy only the first type, fixing what breaks. The expensive surprises usually come from neglecting the last, because deferred patching and unmonitored systems are where small problems become incidents.
What Does Neglected Software Cost an Australian Organisation?
The cost rarely appears as one line in the budget. It shows up as security exposure, slow change and dependence on a few people.
Security exposure is routine, not rare
- The ASD's Annual Cyber Threat Report 2024–25 records over 84,700 cybercrime reports. The average self-reported cost per report was $56,571 for small business, $97,166 for medium business and $202,691 for large business. ASD cautions that the large-business figure may be skewed by outliers and a rise in business email compromise losses, that the data is self-reported and that most cybercrime goes unreported. Treat the figures as indicative.
- The OAIC recorded 1,205 data breach notifications in 2025, up 8% from 1,112 in 2024. Of these, 716 (about 59%) were malicious or criminal attacks. The OAIC's dataset period (1 July to 31 December 2025) differs slightly from the calendar-year headline, so treat the totals as a guide.
Unpatched, unsupported applications are not the only cause of these incidents. They are, however, among the causes an organisation can control.
Slow change and key-person risk
Neglect also appears as:
- Change that takes months. When nobody maintains the application, a small request needs a mini-project to rediscover how the system works.
- Key-person dependency. If one employee or one supplier holds the knowledge, a resignation or a holiday becomes a business risk.
- Workarounds and shadow tools. Staff build their own spreadsheets and side processes, which sit outside security and records controls.
What this means for you: Maintenance is not a cost centre to minimise. It is how you protect an asset that your operations already depend on.
Which Australian Obligations Make Application Maintenance a Governance Issue?
This is general information, not legal advice. Take advice on how these apply to you.
- Privacy Act, APP 11. Organisations covered by the Privacy Act must take reasonable steps to protect personal information. In practice, patching, access control and monitoring of the applications that hold that information are part of the picture.
- Automated decision-making disclosure from 10 December 2026. Organisations that use personal information in programs that make, or substantially assist, decisions that significantly affect people must describe this in their privacy policy. You cannot disclose what you have not mapped, so knowing what your applications do, and who looks after them, matters. Our automated decision-making audit guide sets out a six-step approach.
- APRA-regulated entities. CPS 230 Operational Risk Management took effect on 1 July 2025 and CPS 234 covers information security. Both put weight on third-party service providers and on evidence. Our guide to CPS 230 and CPS 234 audit evidence explains how to capture that evidence once and reuse it.
For all three, a named, documented support arrangement is easier to defend than an informal one.
Application Maintenance & Support Case Studies at a Glance
The projects below come from C9's published case studies and current client relationships. Details are self-reported by C9.
| Industry |
Case study |
Outcome |
What the published detail says |
| Transport and fuel distribution |
Bonney Group |
Long-running managed support |
Managed services "for many years" across the group's technology stacks, including SQL Server, SQL Server Agent and Integration Services, Power BI and ASP.NET Core, with maintenance, support and monitoring |
| Healthcare |
RHealth (ReferHealth) |
Ongoing support of a connected platform |
Middleware between ReferHealth and GP systems using HL7 messaging, RabbitMQ queues and Duo multi-factor authentication, with links to Best Practice, Medical Director and Medical Objects |
Case Study by Outcome 1: What Does Long-Running Managed Support Look Like?
Bonney Group is a major player in Australia's fuel distribution and transport industry. Its challenge was familiar: a group of companies, several technology stacks and a large volume of operational data that needed to be reliable, reportable and secure.
The published case study describes C9 providing managed services across the group for many years. The work covers SQL Server 2016 R2 for business-critical data, SQL Server Agent and Integration Services for scheduled jobs and data movement, Power BI for reporting and ASP.NET Core for web applications, alongside ongoing maintenance, support and monitoring. The page records more than 1,000 hours of work and describes the benefit as keeping the infrastructure stable and secure while the business focused on its operations. That outcome is C9-reported and not independently measured.
What this shows:
- Support is a relationship, not an event. The value comes from a team that already knows how the systems connect and why they were built that way.
- Scheduled work is part of maintenance. Jobs, data loads and reports need watching as much as screens do.
- Needs change over time. Support volumes rise and fall as a client's projects start and finish. That is normal, and a good arrangement flexes with it.
Case Study by Outcome 2: How Do You Keep a Connected Healthcare Platform Running?
ReferHealth (RHealth) set out to replace the fax-and-phone approach to GP-to-specialist referrals with a transparent digital process. That required a backend that could exchange structured referrals and clinical messages with the practice software GPs already use.
C9 built middleware using .NET Core, with HL7 messaging, RabbitMQ queues to deliver messages reliably, and Duo multi-factor authentication to protect access. It connects with Best Practice, Medical Director and Medical Objects, and includes data extraction agents that run on site at GP practices. The published page records more than 600 hours of work, and C9 continues to provide support and maintenance for the platform. [Add a short client testimonial here once approved.]
What this shows:
- Integrations need continuing care. The systems on either side keep updating, so a connection that worked last year can fail quietly this year.
- Security is a maintenance topic. Authentication, access and message handling all need to stay current for a platform that carries confidential information.
- Resilience is designed and then maintained. Queues protect against downtime, but only if someone monitors them.
If your applications also depend on connections to other systems, our guide to custom software integration case studies shows related projects, and our database development case studies cover the data layer beneath them.
A Six-Step Framework for Application Maintenance & Support
This is general good practice, not a description of any one project. You can start it this week, whoever ends up doing the work.
- Inventory. List every application, its owner, its platform, the data it holds and how critical it is. Include spreadsheets with macros and software run by a supplier.
- Baseline. Before changing anything, check the patch level, backups, access rights, known faults and who can currently fix them. This is your starting point and your risk list.
- Prioritise. Rank the gaps by consequence. An unpatched internet-facing system outranks a cosmetic defect. Agree what "urgent" means, in writing.
- Support. Decide how problems are reported, who responds, how they are prioritised and how users are told what is happening. Agree this before the first incident, not during it.
- Improve. Schedule enhancements and preventive work on a regular rhythm, using user feedback and monitoring data. Otherwise, support becomes firefighting.
- Document. Keep architecture notes, deployment steps, credentials handling and decisions current, so that no single person is a point of failure.
A sensible start: apply the framework to one application first. Prove the rhythm, then extend it.
What Should You Measure?
There is no single standard set of measures. The ones below are practical suggestions, not regulatory requirements, and you should agree them with your provider at the start.
| Measure |
Why it matters |
| Time to acknowledge and time to resolve faults, by priority |
Shows whether urgent problems get urgent attention |
| Patch currency |
Shows how long known vulnerabilities stay open |
| Repeat incidents |
Shows whether root causes are being fixed or only symptoms |
| Change lead time |
Shows how quickly a small request becomes a release |
| User-reported satisfaction |
Shows whether support feels responsive to the people who rely on it |
| Knowledge coverage |
Shows how many systems could be supported by someone other than the person who knows them best |
Treat these as conversation starters, and be wary of any provider that promises specific numbers before they have seen your systems.
Five Questions to Ask Any Application Support Provider
- Which of our applications are business-critical, and who is accountable for each? A credible provider will help you draw that line, not assume it.
- How are faults raised, prioritised and communicated? Ask to see the process, not only the promise.
- How are patches and access changes tested and released? Look for a controlled approach rather than a rushed one.
- How will knowledge about our systems be documented and handed over? You should never be dependent on one person, including the provider's.
- What evidence do you have for systems like ours? Ask for examples on your platform, in your type of environment, not a different one.
If you plan to connect AI tools to the same systems, read MCP security questions before connecting AI to your systems first, because support and security obligations grow with every new connection.
What Do These Case Studies Not Show?
A useful article states its limits. The published evidence does not cover:
- Service levels. The pages do not publish uptime, response times or support hours, so do not assume any.
- Measured results. Outcomes are mostly qualitative. Statements about stability and security are C9-reported and not independently verified.
- Comparisons. The pages do not show how an in-house team or another provider would have performed.
- Current arrangements. Support volumes change over time, so a long history does not promise any particular level of service today.
- Further projects. C9's case studies index has a "View More" control, so more projects may exist.
If any of those points matter to your decision, ask for relevant evidence directly before you rely on this article.
Frequently Asked Questions
What do application maintenance & support services include?
They cover fixing defects, applying security patches, improving performance, supporting users, keeping integrations working and adding features as the business changes.
What is the difference between maintenance and support?
Maintenance keeps the software itself healthy through fixes, patches and improvements. Support helps users and responds when something goes wrong. Most arrangements combine both.
What is the difference between application support and development?
Support keeps an existing application running and improving. Development designs and builds new software. The two often overlap, because a support team that knows the system is well placed to extend it.
Can an application built by someone else be supported?
Often, depending on the code, documentation and access available. Start with an assessment of one application to understand the effort and the risks.
Do these case studies prove cost savings or uptime?
No. They describe the work and the relationship. Results are self-reported and not measured here, so define your own success measures at the start.
How do I know if I need ongoing support?
If you cannot name who is accountable for your key applications, or when they were last patched, a review is a sensible first step.
Conclusion
Application support is easier to judge when you can see what it looks like in practice. These projects show a multi-system environment supported over many years, and a connected healthcare platform that needs care because the systems around it keep changing.
The pattern is practical. Find the application that carries the most risk, name who is accountable for it, put a routine around patching, support and improvement, and document what you know. With record breach notifications in Australia, a new privacy transparency rule starting on 10 December 2026 and growing expectations of third-party arrangements, a clear support arrangement is increasingly part of good governance.
What's next?
Choose the one application your business can least afford to lose and apply the six-step framework to it, starting with the inventory and baseline. If you want to test one critical workflow end-to-end before committing to a broader program, our team at C9 can assist with scoping.
Sources and References
Case studies (C9, self-reported)
Australian statistics and regulation
Software maintenance concepts
This article is general information, not legal, security or financial advice. Check current legislation and guidance for your circumstances.
AI Transparency Notice
This article was produced with the assistance of artificial intelligence tools for research and content drafting. All content has been reviewed and approved by C9 prior to publication.