Network transformation is the planned change of an organisation's network from the architecture it grew into over the years to the architecture its business now needs: typically software-defined connectivity across sites (SD-WAN), security built into the network rather than bolted on at the edge, direct and reliable access to cloud services, and operations that can be seen and managed from one place. For Australian organisations it is usually prompted by a lease of an office ending, a move to cloud applications that the old network makes slow, a security review, or a carrier contract coming up for renewal. This guide covers what network transformation involves, what it is for, how to plan it, and how to keep the decision independent of the vendors who will want to sell it.
What network transformation means in practice
Most business networks were built one decision at a time. A head office link here, a branch connection added when the branch opened, a firewall bought after an incident, a wireless upgrade when the old one failed, and a set of carrier contracts that renewed on whatever terms arrived in the inbox. The result works, most days, and costs more than anyone can readily explain.
Network transformation replaces that accumulation with a design. The design starts from what the organisation needs to do (which applications, from which locations, for which people, with what tolerance for downtime) and works back to the connectivity, security and management that deliver it. The components are usually some combination of the following.
SD-WAN and software-defined networking
SD-WAN separates the control of the wide-area network from the physical links underneath it, so traffic can be steered across NBN, fibre, 4G or 5G links by policy rather than by hand. Sites get resilient, lower-cost connectivity and changes are made centrally. Inside the building, software-defined networking does the same for switching and wireless. The practical benefit is that the network becomes something the organisation can change in hours, not weeks, without a technician on site.
Security in the network fabric
Modern designs put security controls in the path of traffic wherever it flows, including to cloud services, rather than only at a head office firewall. That is the idea behind secure access service edge (SASE) and zero trust approaches: every user and device is verified, traffic to cloud applications is inspected without hairpinning through head office, and a branch or a home worker gets the same protection as the corporate office. For Australian organisations this maps directly onto the Essential Eight and the Privacy Act reforms, which expect control over who can reach what.
Cloud connectivity
When the applications live in Microsoft 365, a SaaS finance system and a hosted line-of-business platform, the network's job is to get people to the cloud quickly and safely from wherever they are. Legacy designs that route all traffic through a central firewall make cloud slow and expensive. Transformation puts internet and cloud breakout where the users are, with the security to match.
Visibility and operations
A transformed network reports on itself: which links are up, how applications are performing, where the security events are. That visibility is what lets a small IT team run a large network, and what lets the organisation hold its carriers and service providers to account.
The carrier and contract layer
Underneath every design sits a set of carrier contracts. NBN Enterprise Ethernet and business-grade fibre now reach most Australian business premises, which changes what is available and what it should cost. A network transformation that ignores the contracts leaves the biggest saving on the table.
Why organisations do it
The reasons are business reasons, not technology reasons.
- Cloud applications are slow or unreliable because the network was designed for an on-premises world.
- Adding or moving a site takes too long and costs too much, which slows growth, mergers and relocations.
- Security has become a board question and the current network cannot show who has access to what, or protect people working outside the office.
- Carrier contracts are renewing and nobody independent has tested whether the services and prices still make sense.
- Resilience matters more than it did. A single link failure that takes a site offline is no longer acceptable when every process runs on the network.
- The network costs more than it should, across links, hardware refresh, support contracts and the staff time spent keeping it running.
How to plan a network transformation
Assess what you have and what it costs
A complete inventory of every site, link, device and contract, with the cost and the failure history of each. Most organisations discover services they are paying for and not using, and single points of failure they did not know existed.
Define requirements from the business
Which applications matter, where people work, what downtime costs, what the security obligations are, and what is coming in the next two to three years (a new site, a merger, a system replacement). Requirements written this way survive contact with vendors; requirements written from a product brochure do not.
Design the target state
The architecture that meets those requirements at an affordable cost: the connectivity model, where security sits, how cloud is reached, how it is operated. Designed once, independently, it becomes the benchmark every vendor proposal is measured against.
Test the market
Carriers, SD-WAN and security providers and managed service providers are asked to price against the design, not against the incumbent arrangement. This is where the independent view pays for itself: the organisation negotiates from evidence about what the market offers, not from what its incumbent would prefer it believed.
Stage the migration
Sites and services move in a sequence that keeps the business running: often a pilot site, then the sites with the worst problems, then the rest, with the old network kept alive until each move is proven. Cut-overs are scheduled around the business calendar, not the vendor's.
Operate and review
Once live, the network is measured against the requirements it was built for, and the contracts are reviewed before they renew rather than after.
What it costs, and what it saves
Every organisation is different, so the useful answer is where the money moves. Transformation typically reduces spend on carrier services (by consolidating links, retiring unused ones and repricing at market), on hardware refresh (by replacing site-by-site appliances with a managed platform) and on the staff and contractor time consumed by a fragile network. It adds spend on the SD-WAN or security platform and on the project itself. Whether the net result is a saving depends on how well the current state was understood before the design started, which is another reason the assessment comes first.
Keeping the decision independent
Network transformation is sold hard. Carriers bundle it with contracts, vendors bundle it with platforms, and managed service providers bundle it with the services they would like to run for you afterwards. Each of them can be part of the answer. None of them can be trusted to define the question.
An independent advisor with no carrier relationships and no products designs the target state from the organisation's requirements, tests the market against it and supports a procurement decided on evidence. That is the difference between a network transformation and a large contract renewal with new branding.
Frequently asked questions
What is the difference between network transformation and a network upgrade?
An upgrade replaces components with newer versions of the same design. Transformation changes the design to match what the business now needs, which usually changes the components, the contracts and how the network is run.
Is SD-WAN the same as network transformation?
No. SD-WAN is one component, often a central one. A transformation also covers security, cloud connectivity, operations and the commercial arrangements underneath.
How long does it take?
Assessment and design usually take weeks. Procurement and a staged migration across multiple sites typically run across several months, sequenced around the business calendar.
Do we need a managed service provider?
Sometimes. The decision belongs after the design and the market test, when the organisation knows what it needs run and what that should cost, not before.
Where should we start?
With an independent assessment of the current network and its contracts. Everything else depends on knowing what you have, what it costs and what it is failing to do.
