Columbus pricing snapshot
These are preliminary monthly planning bands, not provider quotes. Read the method note below before using them, because Columbus has no published asking-rate band and these numbers are derived rather than sourced.
| Requirement | Planning range | Budget assumption |
|---|---|---|
| 1 to 2U | $95 to $215/mo | Basic power and internet allocation |
| Quarter or half cabinet | $650 to $1,250/mo | Shared cabinet environment |
| Standard full cabinet | $900 to $1,800/mo | Typical 3 to 5 kW commit |
| 10 to 20 kW high-density cabinet | $1,550 to $4,000/mo | Cooling method must be validated |
| Private cage | $3,600 to $14,500/mo | Four-rack starting profile; density drives spread |
| 250 kW | $135 to $185/kW-mo | Derived secondary-market band, not a published figure |
| 500 kW | $120 to $170/kW-mo | Assumes scale and multiyear term |
| 1 MW | $105 to $155/kW-mo | Contiguous capacity and ramp required |
Method, stated plainly because it matters here: no research house publishes a Columbus asking-rate band. These wholesale rows are set below the published Dallas and Atlanta bands, which are themselves reported as below the national average of $196.25 per kW per month, on a standard secondary-market discount. That is an assumption, not a source. Cabinet rows use 2026 cabinet benchmarks with density and scale assumptions. Taxes, cross-connects, bandwidth and one-time charges are excluded. The published figures we do carry are on the pricing benchmarks page.
In a market with a published band, a quote outside the range is a negotiating point. In Columbus there is no band, so an outlying quote may simply be what the market is. The way to find out is a competitive process, and that is exactly the thing Columbus makes hardest.
Current Columbus availability snapshot
Central Ohio's largest campuses were built for single tenants and will not quote an enterprise requirement. The available market is the multi-tenant layer beside them, and it is smaller than the region's reputation suggests.
| Option | Footprint | Usable power | Cooling support | Estimated delivery | Last verified |
|---|---|---|---|---|---|
| Anonymized cabinet search | 1 to 4 cabinets | 5 to 80 kW | Air or containment; liquid by review | Target 30 to 90 days | Not yet verified |
| Anonymized cage search | 4 to 20 cabinets | 20 to 250 kW | Density-specific engineering review | Target 60 to 180 days | Not yet verified |
| Anonymized suite search | 250 kW to 1 MW | Contracted usable kW to be defined | Air, RDHx or liquid-ready by facility | Target 6 to 24 months | Not yet verified |
What the all-in Columbus cost includes
Cabinet and power
Normalize allocated power, usable IT load, metering and overage language. Ask specifically how large-load tariff changes flow through to you mid-term.
A/B and cooling
Validate dual-feed assumptions, reserved capacity and the cooling approach at your rack density.
Network and cloud access
Add cross-connects, carrier ports, cloud on-ramps and transit. Carrier choice concentrates downtown, so a suburban campus can carry a real network cost.
Install and escalators
Include cages, cabinets, cabling, turn-up, deposits, annual increases and any utility pass-through.
Nearby and strategic alternatives
Cleveland and Cincinnati are the in-state comparisons and both are thinner again. The comparison that actually matters is Chicago, which prices the interconnection Columbus lacks, and Indiana, which is earlier stage and cheaper still. Running Columbus and Chicago in parallel is the only reliable way to see what network depth is costing you.
The real Columbus pricing problem is quote competition
Price in a colocation market is set less by cost than by how many operators will bid. Columbus has a small number of multi-tenant operators that will genuinely quote a mid-sized enterprise requirement, and a very large amount of capacity that will not quote at all.
The practical consequence is that a Columbus process can produce two quotes rather than five, and two quotes is not a market test. That is not an argument against the market. It is an argument for widening the search geographically at the same time, so the Columbus operators know they are being compared to something.
With five bidders, the low quote approximates the market. With two, the low quote approximates whichever operator happened to have a hall to fill. Both can be good outcomes. Only one of them is repeatable, and the difference shows up at renewal rather than at signature.
Power delivery and the AEP question
Central Ohio's utility has served large data center load for over a decade, which is genuinely an advantage: the interconnection process is well understood and the operators know how to work it. It has also absorbed a great deal of that load, which is the constraint.
Large-load tariff structure and queue position have become live commercial issues in the region rather than administrative ones. For a colocation buyer this rarely appears as a refusal. It appears as a delivery date that quietly assumes a substation upgrade, or an attractive rate on a phase whose power is not yet firm.
- Is the power for this hall energized today, under construction, or queued?
- Which substation serves it, and does the campus have a dedicated one?
- What tariff class applies at my size, and is there a large-load rider?
- If a tariff change lands mid-term, who absorbs it?
- If the date slips because of utility work, what does the contract actually say?
A campus with its own dedicated substation solved power at the start. A campus relying on existing distribution is sharing headroom with everything else on that circuit. That distinction is not in any rate card, and it predicts delivery slippage better than anything an operator will tell you directly.
What we see being built right now
Columbus publishes site compliance plans, the engineering approvals that land before a building permit is issued. That gives an earlier read on a project than most jurisdictions offer, and it is the layer we collect.
| Signal | Figure | Why it matters to your quote |
|---|---|---|
| Filings tracked | 10 | City of Columbus site engineering, 2021 to August 2025 |
| Expansion filings on an existing campus | 4 | The fastest space to reach, because power and fiber are already there |
| Concentrated on one campus | 7 of 10 | Continuous expansion at one site over four years |
| Dedicated substation filings | 2 | The strongest delivery-risk signal in the Ohio record |
| Capacity disclosed | None | Ohio site plans carry acreage and square footage, never megawatts |
Collected from City of Columbus engineering records, current to 3 August 2026. We also measured the county building permit layer, all 675,000 rows, and rejected it: it carries no free-text work description, so no data center term matches anything. A site plan proves construction. It never proves that space is available to lease. See what the permit record shows across every market we collect.
Where in the metro you actually land
South Columbus and Rathmell Road
Where the sustained campus development sits, with a dedicated substation and a four-year continuous build record in the filings. Large footprints, priced per kW.
Downtown and the carrier hotels
The interconnection layer, and where cabinet and cage requirements with real network needs belong. A premium to the suburbs, and usually worth it if you use the network.
New Albany and the northeast
The hyperscale corridor. Almost entirely single-tenant and not a colocation option at enterprise scale, whatever the headlines suggest.
Western suburbs
Newer land assembly, longer timelines, thinner carrier choice, and the submarket most exposed to utility upgrade scheduling.
What actually moves your Columbus number
| Variable | Direction | What to ask |
|---|---|---|
| Number of real bidders | The largest single lever in a thin market | Who else can actually serve this requirement here? |
| Tariff pass-through | An open-ended risk if left unbounded | Is there a cap on utility cost increases flowed to me? |
| Substation position | Dedicated beats shared for delivery certainty | Which substation, and does this campus own one? |
| Submarket | Downtown carries a network premium | Am I paying for carrier density or for megawatts? |
| Rack density | Above roughly 20 kW per rack, cooling drives cost | Is the density supported today, and at what premium? |
| Renewal terms | Matters more where switching options are few | What are my renewal caps, and what does leaving cost? |
In a market with few alternatives, the renewal clause is worth more attention than the first-term rate. A good rate with an uncapped renewal in a thin market is a weak position three years out, because your leverage at renewal is whatever competing options exist then. Negotiate the exit before you enjoy the entry.
Get a verified Columbus shortlist, priced against markets that will compete for it.
Send the initial and future power profile, rack density, network needs and required date. We will test actual fit before comparing economics.