How Do You Benchmark Telecom and Internet Costs Across Multiple Locations?

Why Does Comparing Sites Against Each Other Give You the Wrong Answer?

Because every site is priced against different local conditions, and internal comparison bakes those differences into your baseline. Circuit size, access type, building infrastructure, and regional competition all move the price independently of anything the client did.

The size effect alone is large. Analysis of more than 100,000 telecom RFPs and over a million price quotes found median dedicated internet access pricing of $5.09 per Mbps at 100 Mbps compared with $0.2555 per Mbps at 10 Gbps (Lightyear, "2025 State of Connectivity," March 2025). A 20x difference in unit cost between circuit sizes means a small branch will always look expensive per Mbps next to headquarters, and that gap is not evidence of a bad contract.

Access type matters too. The same analysis found that "Type 2" access, where a provider resells another carrier's infrastructure, carried an 18% median markup over direct "Type 1" access on 100 Mbps circuits. A site priced 18% above its peers may simply be off-net for that carrier, which is a sourcing problem rather than a negotiation problem, and it calls for a different fix.

What Should a Multi-Location Benchmark Actually Measure?

Six data points per site, gathered before any comparison happens:

1. Circuit size and technology at each location, including whether it's DIA, broadband, or a private connection
2. Contracted monthly recurring charge, separated from taxes, surcharges, and one-time fees
3. Normalized cost per Mbps per month, calculated from the two above
4. Access type, on-net or off-net for the serving carrier at that address
5. Contract start date, term length, and renewal date, since a five-year-old rate is being compared against a very different market
6. Actual utilization, meaning what the site is really consuming versus what it's paying for

The sixth is where the largest single finding usually appears. Sites frequently carry bandwidth sized for a headcount or an application that changed years ago, and no benchmark catches that if it only looks at price.

How Do You Find the Right Market Comparison?

Compare against current pricing for the same circuit size, in the same region, at the same access type, in the same quarter. Pricing moves, and the direction has changed recently in a way that affects how you advise a client on timing.

Weighted median pricing for 100 Gigabit Ethernet IP transit fell 12% compounded annually from mid-2022 to mid-2025, and the lowest prices on offer in the most competitive markets reached $0.05 per Mbps per month (TeleGeography, September 2025). The distance between that floor and a typical median is itself the argument for benchmarking; a rate can be well off the best available price and still look normal against a national average. That erosion has since slowed. TeleGeography's outlook for 2026 notes that the historical pattern of transport prices dropping 10% to 20% annually has moderated, with hyperscalers now consuming more than 80% of allocated bandwidth on routes such as trans-Atlantic, which limits downward pricing pressure for other buyers (TeleGeography, December 2025).

The practical implication for an advisor: the assumption that waiting produces a better price is weaker in 2026 than it was in 2023. Regional variation is also wide, with year-over-year pricing in the Lightyear data diverging sharply by geography, so a national average is a poor substitute for a regional comparison.

What Does a Realistic Benchmarking Process Look Like Step by Step?

Seven steps, in order:

1. Build the inventory. Every circuit at every location, pulled from invoices rather than from anyone's memory of what was ordered.
2. Separate recurring from non-recurring. Taxes, regulatory fees, and equipment charges distort per-Mbps math and belong in their own review.
3. Normalize to cost per Mbps per month for every site.
4. Segment by circuit size, region, and access type, so like is compared with like.
5. Source live quotes for a representative sample of sites, since a live quote reflects today's negotiating environment in a way a published rate card does not.
6. Flag the outliers against market, and separate the three causes: overpriced contract, oversized circuit, or off-net access.
7. Sequence the fixes by renewal date, because leverage exists in a window rather than continuously.

That last step is what turns a benchmark into savings. A site that is 40% above market with eighteen months left on term is a different action item from one renewing in sixty days.

Which Findings Usually Produce the Largest Savings?

Oversized circuits and off-net access, in that order, ahead of straight rate negotiation. A site paying market rate for three times the bandwidth it uses is a bigger recovery than a site paying 15% above market for the right circuit, and it requires no negotiation at all.

Off-net access is the second. When a benchmark shows a site well above market with no explanation, the cause is often that the incumbent carrier is reselling someone else's last mile at that address. The fix is sourcing a carrier that is on-net in the building, which is exactly where breadth of supplier coverage decides whether an advisor can act on the finding. HonestStok's 4,000+ points of presence exist to answer that question at a specific address rather than in general.



How Should an Advisor Present This to a Client?

Lead with the normalized number and the market comparison, then the recommended action per site with its renewal date attached. Clients respond to a per-site table showing current cost per Mbps, market cost per Mbps, and the variance, because it converts a vague sense of overpaying into a specific list.

Keep the caveats visible. A site above market for a documented reason, such as a genuine redundancy requirement or a building with a single available provider, should be labeled that way rather than counted as recoverable savings. Advisors running this analysis for clients can source the comparison quotes through HonestStok's supplier network at no cost to their client, since HonestStok is compensated by the supplier a client ultimately selects. The FAQ covers how that works, and the about page covers the model behind it.



What's the Bottom Line?

A multi-location telecom benchmark is only useful when every site is normalized to cost per Mbps and compared against current market pricing for that size, region, and access type. Internal comparison finds the worst site; market comparison finds what the estate should cost. Sequence the fixes by renewal date, separate oversized circuits from overpriced ones, and label the sites that are above market for legitimate reasons. Advisors who want market pricing to benchmark a client estate against can work with HonestStok directly or read related posts on the HonestStok blog.