Understanding Egress Billing: Usage-Based, 95th Percentile Billing, & Unmetered

In the ever-evolving world of cloud computing, understanding how data transfer and bandwidth are billed is crucial for businesses.

At ServerCloud, we believe in transparency and value, which is why we’re exploring the differences between usage-based transfer billing and 95th percentile billing.

Usage-Based Transfer vs. 95th Percentile Billing

Usage-Based Transfer (Egress Billed Per Gigabyte): In this model, billing is based on the actual amount of data transferred out of a network. For example, if you have a 1 gigabit Ethernet port, theoretically, it can transfer up to 324 terabytes (TB) of data in a month. However, achieving this would require sustaining a full 1 gigabit per second continuously, which is rarely the case due to varying traffic patterns. For this reason, ServerCloud sizes all cloud compute instances at 12.5 gigabits, or greater.

95th Percentile Billing: This method bills based on peak usage. Your usage is measured every 5 minutes, and at the end of the month, the top 5% of readings (the peaks) are discarded. You’re billed based on the next highest measurement. This method tends to favor those with brief spikes in traffic rather than sustained high usage. This is most often how cloud service providers and Internet Service Providers actually pay for their bandwidth, but is (today) a dated and less-common and less mature way for hosting companies and cloud service providers to bill their customers.

Real-World Data Transfer: Peaks and Valleys

In a typical scenario, where a 1 gigabit Ethernet port is provisioned, you won’t consistently hit that 1,000 megabit per second cap. With normal traffic patterns – highs and lows – the actual data transfer achieved is likely around 225-250 TB in a month. To reach the theoretical max of 324 TB, you would need to maintain a 1 gigabit per second usage continuously, which is highly improbable for most users.

ServerCloud’s Approach: Generosity and Transparency

At ServerCloud, we define the first gigabit of bandwidth for our cloud users as 324 TB. This is a generous allocation, considering that to realistically achieve 324TB in actual data transfer, customers would likely need to have peak traffic of 1.5 gigabits or more. For this reason, ServerCloud provisions all cloud instances with 12.5 gigabits of available throughput, or greater. With VPS (virtual private server) products, ServerCloud provides valuable full rate gigabit connections which are easily upgradable to 10 and 20 gigabit, for always on-budget unlimited usage.

Brad Mitchell, CEO of ServerCloud and a hosting industry veteran with over 24 years of experience, shares his vision: “At ServerCloud, we’re excited to lead our brand into the future by offering more value and transparency than our competitors. By redefining our first gigabit of bandwidth as 324 TB, we’re not just aligning with theoretical limits but also acknowledging real-world usage patterns.”

Conclusion: The ServerCloud Advantage

In summary, while 95th percentile billing can be beneficial for networks with occasional spikes, usage-based billing provides a more straightforward, predictable cost structure, especially for those with consistent traffic patterns. At ServerCloud, our approach to billing and bandwidth allocation is designed to offer the most value and clarity to our customers, embodying our commitment to leading the industry in both innovation and customer satisfaction.