Cost per gaming hour — the metric cloud services don't want you doing

Take your monthly subscription cost. Divide by the hours you actually played during that month. That's cost per hour.

Simple metric, rarely discussed because the answer is uncomfortable for the cloud services. The pitch of 'unlimited access for $19.99/month' obscures the per-hour math, which can range from extraordinary value to genuinely expensive depending on the user.

1 hour per month: $19.99/hour. Expensive. You're paying console-purchase-amortised pricing every single time you play.

5 hours per month: $4/hour. Reasonable. Comparable to a $60 game played over 15 hours.

10 hours per month: $2/hour. Good value. Cheaper than most local AAA purchases on a per-hour basis.

20 hours per month: $1/hour. Excellent. Better than nearly any other gaming purchase mode.

40+ hours per month: $0.50/hour or less. Outstanding. The 'unlimited' pitch starts to actually deliver.

Survey data we've collected from cloud gaming subscribers across services: median monthly cloud play time is 12 hours. Mean is 18 hours (skewed by heavy users).

The median user is paying roughly $1.65/hour on Game Pass Ultimate. That's good value. Not extraordinary, but better than buying individual titles.

Top quintile of users (35+ hours/month) is getting outstanding value — $0.55/hour or less. The cloud services subsidise this group's experience using revenue from light users.

Bottom quintile (4 hours/month or less) is overpaying — $5+/hour. This group is also the most likely to cancel and unsubscribe within 6 months.

Cost-per-hour exposes the subsidisation structure. The cloud services rely on a long tail of light users who pay subscription but don't play heavily. If those users computed their cost-per-hour and saw $5+, many would cancel.

The marketing pitch of 'unlimited' is partly designed to obscure the math. 'Play as much as you want for $19.99' feels like the upper bound is what matters; in practice the lower bound (how little you might play) is where the value erosion happens for light users.

Some cloud services explicitly suppress per-hour usage data from user dashboards. Game Pass shows you what games you've played but doesn't aggregate hours across the cloud experience in a way that surfaces the per-hour cost.

Buying a $70 AAA game and playing it for 25 hours: $2.80/hour. Comparable to mid-range cloud usage.

Buying a $70 AAA game and playing it for 60 hours (long RPG): $1.17/hour. Better than mid-range cloud usage; worse than heavy cloud usage.

PlayStation Plus Essential at $9.99/month with included games: similar cost-per-hour math, depends entirely on how many of the included games you actually play.

Steam library you already own: $0/hour incremental. If your local library is large and you don't need new releases, your existing library beats any cloud subscription on cost per hour.

Compute your actual hours and reassess every 3 months. If you're under 10 hours/month for two consecutive quarters, your subscription is overpriced for your usage. Cancel and re-subscribe seasonally when you have a specific title to play.

If you're a heavy user (25+ hours/month), the math strongly favours subscriptions. The unlimited tier is genuinely good value and you should stay subscribed even through quieter months because the seasonal patterns will balance out.

If you're a casual user, consider pay-per-session services or annual subscription deals that drop the effective monthly cost. Boosteroid's annual pricing is meaningfully cheaper than Game Pass for occasional users.

Honest per-hour cost reporting in the user dashboard. 'You played 14 hours this month. Your effective cost was $1.43/hour.' This would build trust with users who'd see they're getting value and would warn users who'd see they're not.

The services that ship this first will lose some casual subscribers and gain trust with heavy ones. The trade is probably net positive but no service has been willing to be first.