Analzying Spikes in Traffic
Posted by jonathan at 6:03pm EST on 02/10/2009
Over at the Google Analytics blog, they have a column that runs every Tuesday, discussing advanced topics in Google Analytics. Today, was a guest post written by Jessica Hullman of Pure Visibility, an Authorized Google Analytics Consultant.
In her guest post, she discusses some key elements as to why you may see a spike in traffic, and better yet, how you can analyze it effectively determine what caused it.
Here are few things she talks about.
- The number of visits from a certain referrer, such as stumbleupon.com, suddenly jumps when a blog post gets attention
- Your conversion rate falls to zero because of an error in your signup form
- Your percentage of new visitors suddenly increases from an event sponsorship
- Your bounce rate spikes due to a redirect error
Here are a few other key elements to the article that I found relevant.
Useful tips for analyzing traffic spikes with Motion Charts
Focus on cases where significant movement is limited to a smaller number of the total bubbles. Sometimes, multiple bubbles in the chart move significantly. You might find yourself asking, are these spikes equally important? A general guideline you can use to assess the relative importance of spikes is to look at how much the other bubbles in the chart move. If all or most of the bubbles change position drastically throughout the time period, then the spikes probably aren’t very important. If a single bubble moves drastically while the others stay in roughly the same regions, you might have found something interesting.Ask questions to find out what marketing and company events might help explain the trend. Related online (and offline) events are often the cause of traffic spikes from a segment. Keep a detailed schedule of the dates when new advertising campaigns are launched, whether paid search, TV, or print advertising. Keep track of corporate events, news, and press releases, and don’t forget about events that may have impacted the industry at large.
Look to other reports to confirm the trend. In cases where you observe a spike but aren’t yet sure why, dig deeper using additional Google Analytics reports. Often, a spike in traffic from one segment will be reflected by spikes for the same dates in another report. Try comparing Keyword, Geographic Region, Traffic Sources, or Conversion data as you hunt for clues. For example, doing a Date Range comparison of the days directly before and after the spike might show us a similar spike in the traffic from a particular geographic region. Additional information like this might be enough to help explain the cause of the spike.
Head on over to the Google Analytics blog and read what Jessica has to say.
