Nothing Measured, Nothing Gained
Generally speaking, I subscribe to the belief that everything a business puts out on the Internet should be measured in some way. Granted, this can seem a little tedious, and there are times where the logistics that go into collecting data can appear to be complex or time consuming to those responsible for managing the message. In my experience, however, the benefits of being able to make informed decisions of where a company and its people should concentrate resources based on solid analytics pretty much always pays off in the long run.
Take Twitter, for example. Many organizations and professionals have engaged the global conversation that is Twitter in hopes of connecting with potential clients and to bring attention to their own products and services. Certainly, there can be a great deal of benefit in participating in this form of social media networking when implemented well, though blatant advertising only works under very specific circumstances. In fact, anyone using Twitter as a communications tool quickly realizes that it is far more effective to converse in a human sort of way than to blindly attempt to herd folks to the corporate web site. Still, if the end goal of the campaign is ultimately to increase traffic to your web properties, than visits generated by Twitter posts (OK, tweets) is definitely something you want to measure.
You can, of course try and guesstimate traffic coming in from Twitter by having your web analytics package report on traffic coming in from twitter.com, but that solution will omit traffic from stand alone clients (probably the bulk of users these days) and cases where people share the link from your tweet with others via email or some other means.
A far better way is to treat Twitter traffic as you would any other online marketing campaign. And, as most web analytics packages have the ability to track campaign traffic separately from regular web site traffic, you will be able to then track the effect of your Twitter communications as it relates to your web traffic as a whole.
For example, if Google Analytics is your web analytics package of choice, simply create a URL with the campaign tracking parameters using the Google Analytics URL Builder as shown below.

Then, take the generated URL (the big long one at the bottom of the image above, with the utm codes included) and use a URL shortener (another set of utilities that have sprung up as a result of Twitter’s 140 character limit) like Bit.ly or Tinyurl.com to convert it to a version much more suitable for use with Twitter.
Bit.ly Conversion Step 1:

Bit.ly Step 2:

Now, whenever you use the shortened URL in Twitter, you will be able to track and analyze the traffic resulting from those links because they will be grouped in to the Twitter campaign (or whatever name you’ve given the campaign). If your Twitter profile has a link to the web site, you may want to replace it with the Google-generated URL as well, so you can track traffic coming in through that link. The image below shows the results of the campaign tracking in Google Analytics.

From this point you can further refine your analysis based on region, operating systems, etc., as with any other online marketing campaign. And obviously this method of tracking has applications far beyond Twitter. It can, and should, be used for other social media campaigns, email newsletters, etc. It really allows you to get a much better idea of what your audience is interested in, and provides some guidance as to the best ways of engaging social media tools in a much more intelligent way. Good Luck!
Incidentally, I have covered a fair amount of information in this post, and it is a fairly broad example, more to illustrate the point that provide step by step instructions. Anyone interested in learning more about using this technique, or if you’re wondering how to implement something like this within your own environment, feel free to leave a comment, or contact me directly.
As an aside, the jump in Twitter popularity has produced numerous services and tools designed to measure the success or impact of a Twitter campaign. Twitalyzer will evaluate the activity of a Twitter account against a number of social media indicators including influence, signal to noise, generosity, velocity, and clout (see their page for definitions). With Twist you can identify key word trends (your company name, perhaps), and see how they are trending over time. And TweetEffect matches your tweets to your followers in an attempt to give some idea of what messages brought you the most attention. There are a myriad of utilities, and even most of the large Web Analytics vendors are including some level of Twitter analytics tools in their enterprise level offerings. Being as the bulk of this post centered on Twitter, I thought it might be useful to flag some of these for anyone just beginning their Twitter conversation.
Previous post: The Web Intelligence Steering Committee
Next post: Website Performance – The Need For Speed





