Deep-dive guide
Video CDN Cost Simulator: What Streaming Minutes Really Cost Per Viewer
Executive industry analysis
Video is the most expensive thing most products ever deliver. A page of text is a few kilobytes; one hour of high-definition video is several gigabytes. When a streaming business grows, its audience, its watch time and its quality all push the same number upward, and that number is the data you pay to deliver. Because each of those can rise at once, delivery cost often grows faster than the audience itself, while revenue per viewer usually stays flat or grows slowly.
The structure of the problem is a simple chain of conversions. Minutes watched, multiplied by bitrate, give gigabytes per viewer. Gigabytes multiplied by price per gigabyte give delivery cost. Everything else, such as storage, origin traffic on cache misses, request fees and licences, is a smaller layer on top. Revenue per viewer is the counterweight. When cost per viewer approaches revenue per viewer, growth stops adding profit, and every new viewer adds a loss that fixed costs can no longer hide.
The domino model follows that chain in five steps. Data per viewer is set by watch time and bitrate. It sets streaming cost per viewer once priced. Spread against revenue, it becomes the streaming load. After payment fees, what remains is the margin, and the fixed costs of the team and tools decide whether the month is profitable. A small rise in average bitrate, for example when more viewers move to large televisions, can turn the second box amber without anyone changing a price.
The iceberg separates the invoice you watch from the costs you do not. Above the waterline are CDN delivery and storage. Below it are origin egress when the cache misses, request fees for every video segment, DRM and token licences per viewer, and payment processing fees on the revenue itself. Individually they look small, but together they add a meaningful share of cost. Use the simulator to find the watch time, bitrate or price at which your month reaches zero, and then decide whether to encode smarter, renegotiate or reprice.
Formula and methodology blueprint
The model works per month. Data per viewer is minutes watched times sixty seconds times bitrate in megabits per second, divided by eight and by a thousand to give gigabytes. Total data is that figure times the number of viewers. Delivery is data times the CDN price. On top the model adds origin egress for a share of traffic that misses the cache, storage of the whole library, per-segment request fees, DRM per viewer and payment fees on revenue, then subtracts a fixed monthly cost.
- Data per viewer
- GB = minutes × 60 × Mbps ÷ 8 ÷ 1,000
- Converts watch time and bitrate into gigabytes delivered to each viewer.
- Delivery cost
- D = viewers × GB × CDN price
- The main line item, which grows with audience, watch time and bitrate together.
- Origin and requests
- O = total GB × 6% × origin price + viewers × minutes × 15 × request price
- Cache misses fetched from origin and one request for each four-second segment.
- Streaming cost per viewer
- (D + O + storage + DRM) ÷ viewers
- What each active viewer costs you before payment fees and fixed costs.
- Net profit
- Net = revenue − streaming costs − payment fees − fixed
- The monthly result. The breaking point is where this reaches zero.
Cascading elasticity
Watch time and bitrate are the most elastic inputs because they multiply each other and then the audience. Price per gigabyte follows, as it scales the whole delivery line. Revenue per viewer is equally powerful on the income side. Storage and viewer count are gentler: viewers add revenue and cost together, so only the margin per viewer decides whether more audience helps.
Three real-world case studies
Bootstrapped starter
A niche course platform
A small platform serves three thousand learners who watch two hours a month at modest quality. Each viewer pays well, the library is small and delivery is a tiny share of revenue. Load it to see how forgiving video economics are when revenue per viewer is high and watch time is low.
Scaling mid-market
A large catalogue with a negotiated contract
A service reaches one hundred fifty thousand viewers and has secured a low committed price per gigabyte. Delivery is large in absolute terms but well covered. Load it to see how scale and a strong CDN contract keep the margin healthy, and how exposed it would be to a price increase.
Black-swan crisis
Binge viewing on a thin subscription
A cheap plan attracts heavy viewers who stream nearly twelve hours a month at a top quality ladder, while the CDN contract is pay-as-you-go. Revenue per viewer is low and the library keeps growing. Load it to see delivery cost swallow the income in a single month.
Industry benchmark matrix
| Metric | Danger zone | Industry average | Top 10% elite |
|---|---|---|---|
| Streaming cost as share of revenue | Above 45% | 15–30% | Below 12% |
| Data per viewer per month | Above 12 GB | 3–8 GB | Below 3 GB |
| Blended CDN price per GB | Above $0.06 | $0.015–$0.04 | Below $0.01 |
| Cache miss share | Above 15% | 4–8% | Below 3% |
| Margin after streaming and fees | Below 35% | 55–70% | Above 75% |
Variables and levers
Active viewers per month
Distinct people who press play at least once in the month.
Lever: Grow the audience only where revenue per viewer covers the cost, and remove inactive accounts from free streaming.
Minutes watched per viewer
Average watch time per viewer per month, across all titles and devices.
Lever: Use autoplay limits and lower default quality on mobile, because heavy viewers can cost far more than they pay.
Average bitrate
Average video bitrate actually delivered, which depends on the quality ladder and on viewers' connections.
Lever: Adopt per-title encoding and cap the top rung, since many viewers cannot see the difference above a certain bitrate.
CDN price per GB
Blended delivery price per gigabyte after volume tiers and regional differences.
Lever: Negotiate a committed-volume contract or use several CDNs, because price per gigabyte is the biggest lever at scale.
Revenue per viewer
Average monthly revenue from each active viewer: subscriptions, ads and purchases together.
Lever: Raise prices, add ad support or bundle tiers so revenue per viewer keeps pace with watch time.
Library in storage
Total size of all stored renditions, including every quality level and any backups.
Lever: Delete unused renditions and move cold titles to cheaper storage classes, because the ladder multiplies every upload.
How to use this simulator
- 1Enter viewers, minutes watched and average bitrate from your analytics.
- 2Add your blended CDN price per gigabyte and the size of your stored library.
- 3Set revenue per viewer from subscriptions, ads and purchases combined.
- 4Press Find my breaking point to learn which input pushes profit to zero first.
- 5Compare the current set-up with per-title encoding and a committed contract over 24 months.
Mistakes to avoid
- Using the headline CDN price. Regional rates, minimum commitments and overage tiers move the blended price, so use the figure from your last invoice.
- Looking at average watch time only. A small group of heavy viewers can account for most of the data and most of the cost.
- Forgetting that quality settings multiply. A higher top rung raises bitrate, storage and sometimes origin traffic at once.
- Measuring growth by audience. More viewers help only when revenue per viewer exceeds cost per viewer.
Frequently asked questions
Why is video so expensive compared with other content?
Because it is data heavy. Each minute at a typical bitrate is tens of megabytes, and a CDN charges for every gigabyte delivered, so cost grows with minutes, not with page views.
Does a lower bitrate always save money?
It saves delivery cost, but too low a bitrate hurts quality and can reduce retention. Per-title encoding is popular because it lowers the average without a visible drop.
Is storage a major cost?
Usually not, but it multiplies with every quality level and backup. It matters for large libraries that are rarely watched.
What are origin misses?
When a file is not in the CDN's cache, it is fetched from your origin, which can carry its own egress fee. Better cache settings reduce this.
How do I find the breaking point?
Press Find my breaking point. The tool moves each input toward its worse end and reports the value where monthly profit reaches zero.
How accurate is this?
It is an educational model with stated assumptions, not financial advice. Replace the inputs with your own bandwidth reports and invoices and check current vendor pricing.
This is an educational planning model with stated assumptions. It is not financial, tax or legal advice. Replace the inputs with your own figures and verify important decisions with a qualified professional.