Do You Actually Need SD-WAN? A Decision Framework

If you've talked to a network vendor in the last few years, you've been pitched SD-WAN. Sometimes it's the right answer. Often it's overkill — sold to businesses whose actual problems would be solved by simpler, cheaper tools.

Here's the honest framework for figuring out which side of that line you're on.

What SD-WAN actually does

SD-WAN stands for Software-Defined Wide Area Network. The name is technical and unhelpful. Here's the plain version.

A typical business has one internet connection per location. If it goes down, the location is offline. If it gets congested, everything slows. If you have multiple locations, you typically need a VPN to connect them, which adds latency and complexity.

SD-WAN sits between your internet connections and your network. It does four things:

• Combines multiple connections into one logical network. You can have fiber primary + cable secondary, or fiber + cellular, or two fibers from different carriers — and SD-WAN treats them as one resource pool.
• Fails over automatically when something breaks. When your primary connection drops, traffic shifts to the secondary in seconds, often without users noticing. No manual intervention.
• Steers traffic intelligently based on what it is. Voice and video go on the best-quality path. Bulk file transfers go on whichever connection has capacity. Latency-sensitive apps get priority.
• Connects multiple locations as one network. Sites talk to each other directly without going through a central hub. Centralized policy and visibility across the whole network.

The marketing pitch makes it sound revolutionary. The reality is that the value is concentrated in specific situations.

When SD-WAN clearly pays off

Some scenarios where SD-WAN is worth the cost almost every time:

• Multi-location businesses. Five, ten, fifty sites. The connectivity, policy, and visibility benefits scale with the number of sites. Below 3–4 locations, the math gets thinner.
• Voice or video that must not drop. If your phone system or constant video calls go through your internet, automatic failover is the difference between "internet hiccup" and "we lost the customer call." For revenue-critical voice, SD-WAN earns its cost.
• Cloud-heavy operations across locations. If every site is constantly hitting Salesforce, Microsoft 365, Adobe Cloud, AWS, then traffic optimization across sites matters.
• Healthcare, legal, or financial with strict uptime requirements. Compliance often requires the kind of redundancy SD-WAN delivers cleanly.
• Replacing MPLS. If you have an aging MPLS network, SD-WAN typically replaces it at 30–60% lower cost with more capability. This is the most common business case in 2026.
• Real-time logistics, dispatch, or operations. Where minutes of downtime cost real money.

When SD-WAN is probably overkill

Equally important to be honest about: situations where SD-WAN doesn't pay off.

• Single-location small office. Less than 25 people, one location, light cloud usage. A solid fiber connection with a cellular backup router (much simpler than SD-WAN) is usually enough.
• Tolerable downtime. If your business can absorb a couple hours of internet outage per year without measurable damage, the failover sophistication isn't earning much.
• Most work doesn't depend on the internet. Light email, occasional Zoom, basic browsing. SD-WAN's traffic optimization adds little.
• Budget pressure with simpler problems. Sometimes "we just need better internet" is the real answer. SD-WAN on top of bad connections doesn't fix bad connections.

SD-WAN vs the alternatives

A clear comparison helps:

For businesses with just one or two locations and modest reliability needs, the row "manual or no failover" is acceptable. Use a cellular backup router. Done.

For businesses where the row "automatic failover in seconds" is the difference between operating and not operating, SD-WAN is genuinely the right tool.

How it's priced

SD-WAN pricing varies more than other categories. Rough mid-2026 ranges:

• Small site (one user-to-50 users): $150–$400/month per site, plus internet connections
• Mid-size site (50–250 users): $400–$1,200/month per site
• Large or HQ site: $1,000–$5,000+/month
• Cellular failover circuit add-on: typically $50–$150/month

The total is usually structured as a per-site monthly fee plus the underlying internet connections plus optional features like security add-ons (firewall, IDS/IPS).

For a 5-location SMB the all-in monthly SD-WAN cost is typically $1,500–$3,000 plus the underlying connectivity. For a 25-location enterprise, $8,000–$15,000.

Single-vendor SD-WAN vs multi-vendor

Some SD-WAN products are sold by the same carrier that provides your internet ("managed SD-WAN"). Others are sold independently and ride over whatever internet you bring.

Tradeoffs:

• Carrier-managed SD-WAN is simpler — one bill, one support number, often bundled pricing. The downside is you're locked into their internet and their SD-WAN. Switching either means switching both.
• Independent SD-WAN is more flexible — you can use whatever internet you want, change carriers without changing SD-WAN. The downside is more vendors to coordinate.

Most SMBs do well with independent SD-WAN because it preserves flexibility. Larger enterprises often go carrier-managed because the simplicity matters at scale.

The decision framework

Six questions, in order:

• How many locations do you have? One: SD-WAN is probably overkill. Two to three: maybe. Four-plus: probably worth it.
• What happens when internet goes down? If the answer is "we lose money, customers, or compliance posture," SD-WAN pays. If it's "we send people to lunch early," it doesn't.
• Is voice or video business-critical? If yes, automatic failover matters a lot. If no, it matters less.
• Are your internet connections solid? SD-WAN works on top of good connections. Fix bad ones first.
• Are you replacing MPLS? If yes, SD-WAN is almost certainly the right move.
• Do you have IT staff or a partner to manage it? SD-WAN is more complex than basic networking. Either you need internal expertise or a managed service.

If you answered "no" or "doesn't matter" to most of these, you probably don't need SD-WAN. A solid primary internet connection plus a cellular failover router is much cheaper and handles the most common outage scenarios for single sites.

If you answered "yes" to most of these, SD-WAN is likely the right tool.

What we recommend

We work with the major SD-WAN platforms and primarily deploy VeloCloud and BigLeaf for SMB customers. The reasons:

• VeloCloud is mature, broadly supported, and reliable. Good fit for businesses with serious uptime requirements and IT support.
• BigLeaf is built specifically for SMB simplicity. Less granular control than VeloCloud, easier to deploy and manage, particularly good for voice-quality optimization.

Other strong platforms exist (Fortinet Secure SD-WAN, Cato, Versa, Cisco Meraki, etc.). The right one depends on your environment.

Common questions

The takeaway

SD-WAN is genuinely valuable for the right business. For multi-location, voice-critical, or cloud-heavy operations, it does work that simpler tools can't replicate.

It's also one of the most over-pitched products in business networking. Many businesses being sold SD-WAN would be better served by better internet or a basic failover router at a fraction of the cost.

The honest answer for any specific business depends on the questions above. If you're trying to figure out whether SD-WAN is right for you, we can run through the framework together and give you a real recommendation — including "you don't need it" when that's the right answer. Same pricing as going direct, with no incentive to upsell.