|
|
Yesterday’s post on client pains Avinash Kaushik talked about having too many data sources to choose from and Anil Batra addressed the Accuracy of data and Acting on the findings. Continuing the “Identifying & Solving Client Pains” series is Jason Van Orden and Justin Cutroni.
Jason Van Orden is a consultant, author and speaker on podcasting and social media.
Jason’s Thoughts…
How do you measure the size of a podcast’s audience? Whether you’re a content publisher or an advertiser, this is usually the primary metric to track. As a publisher it helps you gauge the success of your show. It allows you to know if you’re connecting with a growing audience. As an advertiser it tells you the reach you’re campaign will have if you sponsor the podcast.
Ideally you want to know how many people actually listen to the content after downloading it. The trouble is that the technology and standards do not yet provide for measuring this. However, a podcaster can still closely estimate the reach of his content by tracking the download and bandwidth stats for his media file.
Simply tracking the number of hits a podcast file receives provides an inflated measurement of audience size because many browsers and aggregators produce several hits for a single download. It’s important to filter out hits by unique individual. This is usually done by removing duplicate IP addresses from the logs. If possible, you also need to filter out hits from known directory bots. Services such as FeedBurner.com and Libsyn.com have developed this sort of download tracking for podcasts.
An alternate method is to measure the amount of data transfer (bandwidth) of a file from the server over a given period of time. By dividing this number by the size of the file, you get an estimate of the number of times the file was downloaded. For example, a 10 MB file that receives 300 GB of data transfer over a month results in a calculation of 30,000 downloads for that month (300 GB/10 MB). This is a measurement of complete downloads since any partial downloads are averaged out in the division.
It should be noted that, while these numbers provide a quantitative measurement of the reach of a podcast, the value of a social medium such as podcasting is in the targeted nature of the audience and the direct, two-way relationship a podcaster has with them. Download and bandwidth numbers should be measured and used in light of this.
Justin Cutroni is Google Analytics Authorized Consultant working for EpikOne
Justin’s thougts…
Our biggest challenge with clients is not about data quality, specific metrics or how we report analysis. Almost every client has questions about these topics and we usually resolve any issues with a brief conversation. The real challenge is helping our clients change the way they do business. Many of our clients think that web analytics is a one or two month project. We stress that web analytics is a business process that should be used to drive growth.
Our goal is not only to ‘do’ web analytics for our clients, but also help the client integrate the web analytics process for the long term. Ultimately it’s about helping our clients change the way they make decisions for their online channel.
So, how do we help clients overcome this problem? We divide our engagements into two phases, a ‘doing’ phase and a ‘teaching’ phase.
During the doing phase we actually do web analytics for the client. This could be measuring the effectiveness of SEO, SEM, email optimization, etc. We analyze data, come up with a hypothesis, test the hypothesis and make recommendations based on the results. Pretty standard analytics stuff. After ‘doing’ web analytics for a few months we usually have a good understanding of the client’s online business process. We take what we’ve learned and start to formulate a plan to help the client integrate the web analytics process into their organization. There are some clients that have no desire change. They are more than happy to outsource analytics and we’re more than happy to help them with that 
But others are eager to bring web analytics in house and build internal resources that can provide online business insight to those who need it.
Very interesting stuff guys. Jason, thanks for teaching me something new and Justin I couldn’t agree more that Web Analytics is a business process and not a few month project.
A question to Jason please about measuring podcast downloads. How often does it happen that Podcasts get picked up and aggregated by other web sites so that they become available for download from places where you don’t have access to stats? Is that an issue or can it be neglected? Something I was always wondering about. Thank you.
“How often does it happen that Podcasts get picked up and aggregated by other web sites so that they become available for download from places where you don’t have access to stats?”
This can happen on occasion. Most sites that aggregate podcast content refer to the original feed and file on your site. In this case the hits still show up in your site stats.
A problem arises when these sites cache the content. This usually causes an uprising amongst podcasters. Often the podcasting community applies pressure to get the portal site to forward requests for content to the originating site.
However, there are cases where this won’t happen. For example, if you put your video podcast on YouTube, you’re limited to the stats they provide.
I recommend putting your content in as many places as you can. When it comes to stats, you’ll have to analyze how this will affect your metrics analysis and decide if you can live with gathering metrics from multiple places or, in some cases, if you can live with the lack of stats from a portal to gain the increased exposure.
I hope that answers the question.