Posts Tagged ‘statistics’

Why do so many people spend so little time on my site?

Tuesday, November 23rd, 2010

Google Analytics can be a great tool for understanding how people use your site, and how to improve your site, however there are some tricks involved in interpreting the data it provides. The “Length of Visit” report is particularly likely to trip people up.

In “Visitors > Visitor Loyalty > Length of Visit” you can see how long people spend visiting your site:

Google Analytics - Length of Visit

The number of 0-10 seconds visits can be quite high and distressing. Why? Because Google Analytics lumps every bounce visitor (those who only view one page of a web site during their visit) in the 0-10 seconds category. Google states:

“In order to capture the length of a visit, Google Analytics tracks the elapsed time between pageviews. The last page of a visit will not be recorded (as there is no subsequent pageview).

Single-page visits are assigned to the 0-10 second category.”

(from: How are the Length of Visit report values calculated?)

Certain sites (eg. blogs, product support sites, tv guides) are more likely to have a high bounce rate because people may want to only get one piece of information at a time. A high bounce rate isn’t always a bad thing. A site may be aimed at on-going engagement so a higher priority may be the number of times a person re-visits (“Visitors > Visitor Loyalty > Loyalty”), or how frequently they come back (“Visitors > Visitor Loyalty > Recency”).

However using “Advanced Segments” you can change the “Length of Visit” report to only show “Non-bounce Visits” (make sure you un-tick the “All Visits” segment before clicking “Apply”) to get a less skewed view of whether people are engaging with your site for a reasonable amount of time.

Google Analytics - Advanced Segments option

Google Analytics - Non-bounce Visits

Google’s in-page help mentions that the “Length of Visit” report can be skewed if people leave their browser windows open, but why doesn’t it mention the 0-10 seconds lumping for bounce visits? Considering how vastly it can skew the perception of activity on your site, I’m surprised that Google doesn’t provide a note about it on the report, or at least on the “About this Report” help section.

Google Analytics - Length of Visit - About this Report

To get a clearer picture on whether people are landing on your site and leaving quicker than you’d like them to, gaAddons is a useful enhancement to the standard Google Analytics code. It’s a code snippet (free for personal or non-commerical use, or a fee for commercial use) that can be added in the same way as your Google Analytics tracking code. You can set a suitable length of time to not consider a visit to be a real bounce. The default is 30 seconds and you can change how it’s reported but if you use the Event Tracking method (“Content > Event Tracking”) you will see something like this:

gaAddons - over 30 seconds report

If you want to take it further, there are calculations that can be done to determine a “real” bounce, or lack of engagement, rate.

Monitoring the buzz

Tuesday, August 25th, 2009

There is a growing amount of discussion happening about understanding the return-on-investment of social media.

eg. What is the value of spending $100k to encourage chatter? Will it generate sales? Just because people are talking about you, how can you tell if that conversation is actually doing more damage than good?

I’ve been reading a number of articles recently, but this one in particular pointed towards research and mathematics so I followed one link to another to compile this list of reads:

Tweeting from digital agencies

Monday, March 30th, 2009

Chris Lake has tried to track down the twitter accounts of the top 50 UK interactive agencies (as ranked by New Media Age) in his Econsultancy post Digital agencies on Twitter (or are they?). While looking at who is practicing what they preach he found that of the top 50, a couple of agencies are very active on twitter, some slightly active, others have only claimed their names, while many are no where to be seen.

He suggests “roughly a 25% adoption rate” so I decided to spin some numbers for a closer look and found that if you include those who have at least claimed their names then this boosts the figure to 36%, or 18 out of the 50 agencies.

Out of those 18 agencies with twitter accounts, only 33% are independently owned – this surprised me as I had assumed that independent agencies may be more inclined and less restricted by red-tape to leap on to twitter. On the other hand I did have a correct hunch when I found that 67% of the twittering agencies, whether independent or otherwise, have less than 100 staff members – perhaps smaller companies find it easier to react to new technologies and co-ordinate engagement.

Breaking down the types of agencies that are twittering: 67% classify as Marketing, 22% as Design & Build, and 11% as Technical.

The comments to Chris’ post vary from minor debate about whether or not to judge an agency on its twitter presence to pointing out that individuals within the companies are often active tweeters even if there isn’t a company presence. Some responses question the point behind setting up a company account.

Claiming your company brand is useful to avoid brand-squatters and communicating via twitter can help to spread news. A central account can provide a link between staff, helping employees to connect, and potential employees to explore further in to the agency. Depending on the type of agency, a company twitter account can be used to:

  • Share information and findings
  • Seek advice on solving issues
  • Update industry friends or inquisitives about latest campaigns
  • Link to useful articles
  • Develop partnerships with local businesses
  • Promote your business
  • Encourage staff knowledge-sharing
  • Provide an accessible point of contact
  • Announce job vacancies
  • Demonstrate the culture of the company to potential employees
  • Access a community willing to give you a few minutes for testing/feedback
  • Spread the word about industry events
  • Have a laugh

It doesn’t hurt to think about defining some measurements (they might just help you with your next social media pitch) but remember that not all metrics are equal. If you’re interested in measuring twitter success, I’d recommend reading yongfook‘s recent post, Social Media ROI, which provides helpful tips and examples of defining metrics.