5G Between Hype and Reality: What the Standard Promised, What Got Built

01 The Promise

When 5G was standardized, its marketing promised three things at once: multi-gigabit peak throughput, single-digit-millisecond latency, and a network soft enough to be carved into virtual slices for factories, vehicles, and hospitals. The ITU's IMT-2020 requirements made those numbers official, and carrier advertising compressed them into a single shorthand: 5G would be not just faster but categorically different from 4G.

A decade into deployment, the honest scorecard is mixed in an instructive way. Coverage is broad, mid-band capacity is real, and average speeds have risen measurably over the 4G baseline. Peak rates, sub-millisecond end-to-end latency, and slicing remain confined to specific deployments rather than general subscriber experience, and the populations seeing each vary enormously by country.

The gap is not primarily a story about radio engineering failure. The radio mostly did what its designers claimed. It is a story about what each promise cost to deliver, and about who was actually willing to pay that price — a question the promotional material largely skipped.

02 What Frequency Buys and What It Costs

Radio spectrum obeys a tradeoff that no standard can legislate away. Higher frequencies offer more bandwidth — wider channels carry more bits — but they propagate shorter distances, penetrate walls poorly, and require more equipment per square kilometer. Lower frequencies travel far and penetrate buildings well, but their narrow channels limit the capacity any single site can deliver.

This is why 5G describes at least three quite different networks wearing one name. A low-band layer at 600 MHz delivers coverage that lights up a 5G icon on a phone while performing closer to good 4G. Mid-band spectrum near 3.5 GHz carries the bulk of the real-world capacity gains. Millimeter wave delivers spectacular speeds within line of sight of an antenna.

None of these choices is wrong. Each buys a different point on the same physics curve, and deployment economics decide which points get funded first.

03 The Physics of mmWave

Millimeter-wave spectrum — roughly 24 GHz and above in 5G allocations — is where the ambition met the atmosphere. Signals at 28 GHz attenuate sharply in air, are blocked by foliage, walls, and even energy-efficient glass, and diffract poorly around obstacles, which is a polite way of saying that a person standing in the path matters.

Compensating is expensive. Because antenna elements scale with wavelength, mmWave bands allow arrays of hundreds of antennas in a small panel, and beamforming focuses energy toward each individual user. That restores the link budget on paper, but it means base stations must effectively see the user, and cell handoffs become frequent at anything faster than walking pace.

The result is genuine where it is built — stadium seats, dense urban corridors, fixed wireless into a rooftop dish — and largely absent everywhere else. Physics did not fail; it priced the service.

04 The Mid-Band Compromise

The center of gravity of real 5G sits at 3.5 GHz-class mid-band spectrum, and most of the measurable improvement in consumer experience comes from this single layer. It balances range measured in hundreds of meters with channels wide enough to deliver several hundred megabits per second in good conditions, while massive MIMO antenna arrays multiply capacity on existing tower grids.

Mid-band also exposed a policy problem. The United States fragmented its mid-band holdings among incumbent users and auctioned additional spectrum late, while much of Asia and Europe aligned on the 3.5 GHz range earlier and built on it faster. The radio engineering was essentially identical; the regulatory timelines were not.

When national 5G outcomes are compared, spectrum policy explains more of the variance than vendor choice or radio innovation. That is an uncomfortable conclusion for an industry that prefers to market silicon and speed.

05 What Actually Got Built

Sub-6 GHz — overwhelmingly a mid-band capacity layer plus low-band coverage — is what carriers actually deployed, and by coverage metrics it is a genuine success: a large majority of the world's population is within reach of a 5G-capable signal. mmWave remains a small fraction of deployed sites, concentrated in dense venues rather than spread across geography, and several major carriers have quietly deprioritized it entirely.

The investment line tells the same story. Operators worldwide committed enormous sums to spectrum auctions and site upgrades, and payback has run slower than the 4G cycle because consumers proved unwilling to pay a premium specifically for 5G. Revenue gains arrived mostly through traffic growth rather than a 5G price tier.

The build-out, in other words, followed return on investment, and ROI consistently favored the compromise layer over the flagship one. Anyone forecasting the next radio generation should start from that observed preference rather than from the launch-day headline rates.

06 Latency and Slicing in Practice

Latency was the promise with the widest gap between laboratory and street. IMT-2020 targeted 1 millisecond for ultra-reliable use cases; deployed networks deliver radio-layer latency in the tens of milliseconds once scheduling, backhaul, and core processing are counted. The radio interface did improve measurably over 4G, but the end-to-end path is dominated by transport and application components the standard only partially governs.

Network slicing — partitioning one physical network into isolated virtual networks with guaranteed characteristics — exists in the standards and in early commercial offerings, mostly for private industrial deployments such as ports and factories. It is genuinely useful where enterprises pay for it. It has not reached general consumer service, because ordinary subscribers do not currently pay for guarantees.

Both gaps share a shape: the radio standard delivered what it said, and the economics of the rest of the system did not follow it. That distinction matters when assigning blame for unmet expectations.

07 The 6G Question

Discussion of 6G is already repeating the pattern: headline numbers first, deployment economics later, with terahertz frequencies and integrated sensing proposed as the new frontiers. The lesson the 5G decade actually teaches is to read those numbers as capability envelopes rather than delivery commitments.

What would make 6G different? A deployment story in which the expensive layer has a paying customer attached from day one — fixed wireless, industrial automation, or sensing services — rather than a consumer upgrade market assumed by analogy with the 4G era. Early operator commentary suggests exactly that reframing is underway.

The underlying physics will not change between now and 2030. Spectrum will still trade capacity against range, and the networks that get built will still be the ones whose bills someone agrees to pay. That is not cynicism; it is the summary of one complete deployment cycle, drawn from a decade of observed evidence.