Valitics Blog

One of the core goals of almost all websites (for good or bad) is to get people to come back once they have visited.  This generally makes sense because the cost of retaining should be less than the cost of initial contact.  I thought I would give you a couple of metrics for determining whether your website is achieving the goal of visitors "coming back for more."  Below are a few of the concepts related to getting people back, and a potential metric (or two).

Great Content

Great content is probably the best way to get people to return to your site.  I have a few metrics for measuring great content, and you can't really use any of them in isolation and come up with a reasonable assumption about your content.  There is no way to use a single metric with no analysis to determine the quality of your content.

New to Return Visitor Ratio -- This will give you an idea of your customer mix.  If one of your goals is to have more return visits this ratio should be high.  However, because it is a ratio, if you work hard to get new visitors and you work on retention, the scales may tip.  As a general rule you want the ratio to reflect how you allocate your marketing resources.  If it doesn't you may want to reevaluate the type of marketing/advertising you are doing.

Number of Times Visited -- This will give you a clearer view of website loyalty than New to Returning Ratio.  Again, be careful about making assumptions without context (competitors may visit your site many times, but that doesn't increase the propensity to be a loyal customer).  Depending on your analytics this measure may be lifetime visits or period visits.  If it is lifetime (like Google Analytics) it is difficult to pin retention to any specific campaign, but you can see the likelihood that people will return.

Top Content -- Look at the top pages of the site ranked by unique pageviews and compare that to total pageviews; you might be surprised that your home page is not always at the top of the heap.  Visit the top pages on your website and make an educated assumption about the quality of their content.  If pageviews is significantly higher than unique pageviews, the time on site is relatively high, and you feel the content is compelling you are getting close to safely saying the page has good content.  Keep in mind, the purpose of the page greatly affects your assumption.

Newsletters

Newsletters are a good way to bring people back to your site.  If folks double-opt-in to your email campaign you are almost guaranteed at least one return visit (don't mess it up :-).  What you want to find out is, what is it that makes people come back for more.  There are a number of metrics you could use for this, but my favorite is:

Campaign Impressions to Unique Campaign Visits  Ratio -- This will take a little calculating, depending on your analytics solution.  Basically, it is a measure of the number of emails opened (an "opportunity to view") divided by how many people visited the site (don't count bounce visits or multiple visits in a time period if you can help it).  Once you have this number start comparing it campaign to campaign.  After some comparison you will be able to determine which content, design, and frequency keeps them coming back.  If you have website conversions add that to the mix for gut-checking your newsletters' efficacy.

Forums or FAQ's

If your site serves any sort of support function return visits can be good and/or bad.  You can't make that determination using web analytics, but you can get closer to understanding why people come back.  The two best, non-web analytic, ways to know if your forum or FAQ is working is to (1) notice a decrease in support calls or emails and/or (2) ask your visitors via surveys.  But, to start to get a preliminary look try the following metric and analysis:

Time on Forum Pages -- This is a very deceptive metric, so don't take it at face value.  A bigger number is not always better.  More time on forum pages could mean people are engaged and finding answers; it could also mean they are lost and can't find what they need in an acceptable time frame.  Start by looking at the content on high time pages -- Are there comments multiple comments or strings?  Is there a low bounce rate and high exit rate?  Are there high search engine entrances?  These things likely indicate the content is helping. 

As with all metrics using them outside of appropriate context is dangerous.  You will certainly need to analyze your analytic data, or you are likely to make the wrong decision.