Have you been looking at your network charges and wondering just what 95th percentile billing means? Then wonder no longer. Here is an easy to understand guide that will explain it all.
95th percentile billing, sometimes referred to as Burstable Billing, is an industry standard method of billing for bandwidth usage on a network connection.
All data transferred over the network is sampled throughout the month at regular intervals, say every 5 minutes. At the end of the month the levels are then listed from highest to lowest, the top 5 per cent of the sample is discarded, and the bill is calculated on the highest level remaining – the 95th percentile. For example, to simplify things, if over the course of the month we only took 100 sample readings, and listed them from highest to lowest, they might appear as such:
100 = 8Mbps
99 = 7Mbps
98 = 6Mbps
97 = 5Mbps
96 = 4Mbps
95 = 2Mbps
94 = 1Mbps
The top 5 per cent of those readings would be discarded, and you’d then be charged for 2Mbps as, for 95 of the readings, you were at or below 2Mbps.
Unexpected spikes occur, at one time or another, on almost every website. A particularly popular blog post, a new product release, or successful marketing campaign can lead to sudden, temporary spikes in traffic. When these spikes come out of the blue, they could potentially throttle your bandwidth speeds and incur extra cost.
By using the 95th percentile billing method we are able to both accurately scale up our network capacity in line with peak demands, and ensure our customers are billed fairly without being penalised for brief spikes in network traffic.
This gives you peace of mind that sudden spikes in traffic won’t slow you down and inhibit other critical business functions. And the extra five per cent is not chargeable, so you’re essentially getting that usage for free.
By using the 95th percentile method, we are able to accurately calculate the network’s capacity, ensuring there’s no disruption to the service, poor latency or any other bandwidth complications.
This billing model is widely used throughout the industry for business customers with larger amounts of internet traffic, rather than the standard domestic billing method that charges customers for all data transmitted. If used for business customers with high data demands, this billing method would prove very costly and wouldn’t guarantee the extra capacity for traffic spikes.
Bandwidth caps can be put in place, ensuring that once a certain level has been reached, the supply will not be able to exceed this amount. Whilst it guarantees you’re not able to go over a set level, the capping will have the effect of throttling your bandwidth once that level has been reached, slowing your speeds and restricting your service should you hit any unexpected spikes.
Hopefully that’s explained it all. But if you still have queries, or want to ask a question about any service or billing issue, our dedicated team are on hand to answer your questions.
What other colocation conundrums have you come across? Leave a comment and let us know.
You must be logged in to post a comment.