Getting Started with SD-WAN
SD-WAN (Software-Defined Wide Area Network) is changing how UK businesses connect their offices, branches and cloud services. Instead of paying for one rigid private circuit per site, SD-WAN combines the connections you already have — leased lines, business broadband and 4G/5G — and decides in software which path each application should take. The result is a network that is more flexible, more resilient and usually cheaper to run than a traditional WAN.
Why traditional WANs are struggling
Most multi-site networks were designed when applications lived in a head-office server room. MPLS circuits linked every branch back to the centre, and that model worked well for years. Today, though, much of the busiest traffic goes to Microsoft 365, Salesforce, AWS and Azure. Sending it back through head office before it reaches the internet adds latency, uses up expensive private bandwidth and creates a single point of failure. Adding a new site or upgrading a circuit can also take months.
How SD-WAN works
SD-WAN separates the intelligence of the network from the physical links underneath it. A small edge device at each site connects to every available circuit and continuously measures its latency, jitter and packet loss. A central controller holds your policies — which applications matter most, which links they should prefer and what should happen when a link degrades — and pushes them to every site at once.
Because each application is identified and routed individually, a Teams call can take the cleanest, lowest-latency path while a large software update uses spare broadband capacity. If a link fails or its quality drops below an agreed threshold, traffic moves to another link in under a second, usually without users noticing.
The business benefits
- Flexibility: new sites can be brought online with zero-touch provisioning. A pre-configured device is shipped to site, plugged in and automatically joins the network. Extra capacity means adding a circuit, not re-engineering the WAN.
- Lower costs: by mixing dedicated leased lines with lower-cost broadband, most organisations replacing MPLS with SD-WAN cut their WAN spend by around 25–35%.
- Better application performance: real-time path selection and traffic prioritisation keep voice, video and business-critical systems responsive when a link is busy. Cloud traffic can go straight to the internet from each site instead of travelling via head office.
- Centralised management: every site, link and policy is visible from a single dashboard, so changes that once needed an engineer on site can be made centrally in minutes.
Getting started: four steps
A successful SD-WAN project starts with understanding your current network, not with choosing hardware.
- Audit your traffic. Find out which applications each site uses, where that traffic goes and which systems cannot tolerate downtime.
- Choose the right underlay. Decide which sites need a dedicated leased line, which can run on broadband and where 4G/5G backup makes sense.
- Pilot at a few sites. Prove your policies and failover behaviour on a small scale before rolling out further.
- Roll out and manage. Migrate the remaining sites in phases, with 24/7 monitoring in place from day one.
Many businesses choose a managed service at this point. It gives them the benefits of SD-WAN without having to build in-house expertise in policy design, carrier management and round-the-clock monitoring.
Key takeaways
- Enhanced network flexibility and performance
- Significant cost savings on bandwidth
- Improved application performance
- Centralised management and control
Find out how we deliver this as a fully managed service on our Managed SD-WAN page.
Deciding what to run SD-WAN over? Our guide to SD-WAN vs leased lines explains why most enterprise networks use both.