RSS Feed

Podcasting Metrics Series Pt 5 | Bandwidth as a Podcast Metric

December 8, 2006 by Jason

Categories: Metrics

In Part 4 of the metrics series I talked about how you can measure complete downloads and whether it was necessary. Moving on to Part 5, I’ll take a look at how to keep your site from being shutdown by your host due to excessive bandwidth use and how to measure complete downloads using bandwidth.

Measuring Bandwidth to Avoid Hosting Problems

The obvious reason to measure bandwidth is that it’s the primary cost that you’ll incur as a podcaster. The large size and quantity of the content podcasters deliver add up quickly.

You need to watch your bandwidth usage in order to avoid problems with getting your hosting temporarily, or permanently, shut down. I’ve experienced the growing pains that result in emails from my web host warning that my site (and hence show) will be shut down at any moment because I’m using more than my fair share of bandwidth. That’s no fun when you’ve just released a new episode.

Before you start a podcast, do bandwidth projections to choose the best media hosting strategy for your expected audience size and growth. Then watch your bandwidth usage on a regular basis so you can see when you’ll need to upgrade your hosting infrastructure. If you need help in calculating/projecting your bandwidth needs, take a look at this article on podcast hosting from my How to Podcast Tutorial site.

Using Bandwidth to Measure Your Audience

How does bandwidth do for measuring audience size? It depends on what other information you have to relate it to.

It may be seem impressive when a podcaster brags about two terabytes of data transfer last month, but this number alone is insufficient to draw any conclusions about the size of his audience. Bandwidth doesn’t tell you anything if you don’t also know the frequency with which episodes are released and the average size of each file. With this additional information you can divide the bandwidth metric by the number of shows in the month and the average file size to measure the average audience per episode that month.

For example:

2 TB transfer per month / 4 shows per month = 0.5 TB per show

0.5 TB per show / 20 MB avg file size per episode = 25,000 avg downloads per episode

A more accurate method is to take the amount of bandwidth data transfer (bandwidth used) for a specific file over a given period of time and divide it by the size of that file to calculate the number of downloads.

For example:

Given a 25MB file downloaded over a month…

300 GB of transfer over one month / 25 MB = 12,000 downloads

This method averages out the partial downloads and give you an estimated number of complete downloads. This is another method used by Marsall Sponder to track the podcasting metrics for IBM.

Do you watch your bandwidth stats? How do you use bandwidth as a metric?

Comments are closed.