The New Enterprise

For thirty years, “enterprise-grade” meant feature depth. The biggest routing tables, the deepest packet inspection, the most line cards in the core switch. That definition made sense when a single corporate building held ten thousand people and the servers were down the hall.

That building is mostly empty now. The servers moved to the cloud. The people moved everywhere. And here is the part the industry still hasn’t caught up with: the enterprise did not get smaller. It got more distributed. The requirement moved from feature depth at a central core to resilience, visibility, and manageability across a multiplying number of edges.

I’ve been calling this the New Enterprise, and I’ve watched it arrive for nearly two decades. In 2008, during the financial crisis, an environmental research company needed to move terabytes of wind-farm field data over 3G. The traditional quote was about $250,000. We built it with early multi-WAN bonding for about $25,000 — and it ran better. That was the day the math inverted. It has never inverted back.

The test that matters

Here is the cleanest test of whether a network is truly enterprise-grade, and it has nothing to do with price: on consumer equipment, when something breaks, you reboot it. On real enterprise equipment, you don’t — because when a component fails, the failure is invisible to the user. Rebooting is not a plan. It’s an admission that the architecture depends on hope.

The New Enterprise refuses that bargain — not by buying a more expensive single connection, but by refusing to depend on any single connection at all. Multiple diverse paths stay active. Traffic is distributed at the packet level. When one path degrades, the session continues on the others and nobody notices. That’s the difference between “we have failover” and “we do not go down.”

Who this is for

Maybe five percent of organizations genuinely need carrier-grade cores and datacenter fabrics — they run their own ISP, or they move massive machine-to-machine traffic on premises. The other ninety-five percent were sold that architecture anyway, and they’re paying for feature depth their workloads left behind years ago. The organizations already living the alternative are not hypothetical: cruise ships bonding a dozen or more Starlink terminals into 30+ gigabits, live golf and Formula 1 broadcast running without satellite trucks, retail chains running 1,900+ locations on bonded 5G with support tickets approaching zero.

I wrote the full argument as a white paper — the history, the four pillars, the economics, and an honest section on who still needs the old model. It’s free, and it’s the intellectual backbone of everything I build: The New Enterprise, at Connectivity 101.

The enterprise didn’t shrink. It changed shape. Build for the shape it has now.