Most outage post-mortems we read at Buckeye Telecom start with the same sentence: "Our IT team did everything right." And in almost every case, that's true. The problem isn't the IT team. It's the vendor stack underneath the carrier — the four or five layers of suppliers, sub-suppliers, and platform dependencies that the IT team doesn't even know exists.
When you sign a contract with a national carrier, you're not really signing with one company. You're signing with a chain of them. Most Ohio business owners couldn't draw a diagram of that chain if you handed them a whiteboard. Neither could most of their carrier reps. That gap is where the failures happen.
What's Actually in Your "Managed Service"
Here's what a typical Ohio multi-location business is buying when they sign a managed service agreement with a national carrier.
Layer 1 — The carrier's brand. This is who you write the check to. Their logo is on the bill. Their account team is who you call when something breaks.
Layer 2 — Fulfillment partners. The carrier doesn't actually own all the fiber, towers, or last-mile equipment delivering your service in every market. They lease capacity from regional ISPs, partner with backbone providers, and hand off chunks of your network to third-party MSPs they've contracted with.
Layer 3 — Hosting and cloud infrastructure. Your phone system might be hosted on Amazon Web Services. Your SD-WAN orchestration runs on a third-party SaaS. Your call recording lives on someone else's storage entirely.
Layer 4 — Hardware suppliers and firmware vendors. The Cisco, Adtran, or Fortinet box on your wall has firmware that gets pushed by another vendor entirely. When that vendor pushes a bad update, the box stops working — and your carrier's tier-1 support team doesn't have visibility into it.
Layer 5 — Backbone and transit providers. The connection between your carrier and the rest of the internet rides on backbone networks like Lumen, Cogent, and Zayo. When one of those backbones has a routing problem, your traffic finds a different path — assuming a different path exists.
Any one of these five layers can take you down. None of them have your phone number.
Why This Is Getting Worse, Not Better
Three things are making the vendor stack longer and more fragile every year.
Consolidation in the carrier space. Mergers and acquisitions in telecom have accelerated. When carriers buy each other, they don't unify their vendor stacks — they layer them. A multi-location business in Cleveland might have services that touch four different acquired networks under the same brand name, with no single team responsible for end-to-end performance.
SaaS dependency creep. Modern voice, network, and security services now ride on SaaS platforms that themselves depend on other SaaS platforms. When one of those upstream platforms has a bad day, the impact cascades through the stack with no warning.
Opaque subprocessors. Most carrier contracts allow the carrier to substitute subprocessors without notice. Your service might be delivered through a different last-mile partner this month than it was last month, and you would not know unless you knew exactly what to look for.
The Numbers from the Past Month
April 2026 was not a quiet month for network reliability. The week of April 20–26 saw 264 global network outage events tracked by major monitoring services — 170 of those were in the United States. That's a 19 percent jump from the prior week, and it's sitting on top of a year-over-year baseline that's been ticking up steadily for three years.
A few specific incidents worth noting:
January 16, 2026. Madgenius, a U.S.-based hosting provider with major nodes in Columbus, Ohio, went dark for one hour and sixteen minutes. Multiple downstream partners across multiple regions were affected. None of those downstream customers had a direct contract with Madgenius. Most of them had no idea Madgenius was in their stack at all.
February 20, 2026. A wildlife-related fiber cut took out phone and internet for Medina County government facilities. The county initially blamed contractors. The actual cause was a squirrel that had built a nest inside the fiber infrastructure and chewed through multiple lines.
April 22, 2026. Verizon Business experienced an outage affecting customers across multiple U.S. regions for one hour and three minutes. Direct customers got an apology email. Indirect customers — businesses whose carrier was reselling Verizon Business capacity — got nothing.
The Madgenius and Verizon Business incidents share a common characteristic: the impact was disproportionate to the cause. A small failure at one node took down dozens of unrelated businesses. That's the vendor stack at work.
The "One Number to Call" Test
Here's the simplest way to evaluate any telecom or IT vendor for your business: when something breaks, how many degrees of separation are between you and the person who can fix it?
For most Ohio businesses on a national carrier contract, the answer is somewhere between three and five degrees. You call the carrier's general support line. They escalate to a tier-2 team. That team opens a ticket with a fulfillment partner. The fulfillment partner contacts the upstream provider. By the time anyone with authority is looking at your problem, an hour has gone by and three different ticket numbers exist.
For a business working with a regional partner that owns its supplier relationships, the answer is one. You call a person. That person calls the supplier. The supplier responds because the relationship is real and the volume matters.
That difference shows up most clearly during outages. But it shows up in normal operations too — in change requests, in moves and adds and changes, in billing disputes, in capacity upgrades.
What Actually Works
We design networks for Ohio multi-location businesses every week. Here's what we've learned about keeping the vendor stack short and accountable.
Pick suppliers you can put a face on. Every layer of your stack should have a person attached. If your carrier can't tell you the name of the engineer responsible for your service, that's a layer with no accountability.
Demand visibility into the stack. A good vendor will tell you exactly which subprocessors are involved in your service. A great one will give you incident notifications when those subprocessors have issues — even if your specific service wasn't directly affected.
Build redundancy at the layer that matters. Most businesses think about redundancy at the circuit level: two ISPs to two routers. That's necessary but not sufficient. You also need diversity at the upstream layers — different backbone providers, different hosting platforms, different DNS providers. A redundant circuit that terminates at the same upstream peering point is not redundant.
Insist on direct relationships where they matter. For your phone system, your security platform, and your network monitoring, the vendor should be in the room with you when you design it. Anything sold through a reseller layer that the original vendor doesn't even know about is a layer you can't trust.
Audit your stack quarterly. Vendor stacks shift over time. Subprocessors get added. Service paths change. A quarterly review with your provider keeps surprises from turning into outages.
Contract Language That Matters
When you renegotiate your carrier contract, two clauses are worth fighting for.
Subprocessor disclosure. Require the carrier to maintain and share a current list of subprocessors involved in delivering your service. Require notification — not just consent — when subprocessors change.
Service-credit visibility into upstream events. Many service-level agreements only credit you for outages caused directly by the carrier. If an upstream backbone goes down, you're often on your own. Negotiate for credits tied to actual unavailability, regardless of cause.
Both clauses are negotiable for any business spending more than a few thousand dollars a month on telecom. Most carriers won't volunteer them — but they will agree to them when asked.
Bottom Line
Your IT team is not the problem. The vendor stack they were handed is.
If your network has had more bad days than you can explain, ask your carrier for a complete subprocessor list. Look at the chain. Count the layers between you and the person who can actually fix something. Then decide whether that's the partnership you want for the next five years.
Fix the stack.
Get a Free Assessment from Buckeye Telecom →