Fun with Web Metrics

What's going on?

Understanding Your User Base

So, you’ve got some numbers – say monthly visits.   If your site is just starting out, you’re probably relying on viral exposure to increase traffic: i.e., if you build it, they will come. Or perhaps you’ve done some degree of  marketing, even if it’s only getting your site listed somewhere or noticed by search engines.

If you’ve been around for a while, you might have a regular client base (in that you know that some of your visitors are repeat visitors) though you’re also relying on word-of-mouth from those satisfied users to drive traffic.

So, what does something like “monthly visits” MEAN?

First, let’s assume that every visit to your site is relevant, and optimistically assume that every visitor is absolutely wild about your site.    The question is, if you had the most AWESOME and perfect site on the WWW, such that any potential visitor came there, how many would that be?   What’s your ultimate target population?

For example, if you’re Google, then that’s really the population of the planet that uses the Web.   Take 7 billion people, multiply by the fraction that are web users these days, and you have your answer.  Doing a quick search, sets the web-using population at about 25.6% of the world population, or about 1.7 billion people. Your site probably will be a fraction of that.   🙂

Map of worldwide internet users - 2008

If your site is for a group like the South Dakota Butterfly Enthusiasts it might only be a couple of hundred people.

So – two questions whose answers you should have at your fingertips:

  1. Who is my user base?  (e.g., “anyone who might want to buy one of my custom-built birdhouses”)
  2. How many of them are there out there?

Now you can start worrying about the fun stuff.  🙂   How many of your actual visits are people who fit into your user base?    Consider the links above.   The first one (for those of you who clicked it), is pretty much on target (for them) – you were curious about information and they provided it.   Maybe you even bookmarked it in case those stats might be helpful in the future.   The second link – while a good example of a discreet population of web users – aren’t probably expecting traffic to that page outside of a very select group of people.  Nonetheless, it’s a nice page, has useful information, but now has some percentage of its traffic driven by a random effect (that is to say – you).

Here’s where things get a little complicated and require more digging.   Neither of those pages are the main page for the site.   Most of the traffic from “here” are probably single-page views with a very-short time on site.   However, in terms of visit success it can be argued that visits to both links from here are successful albeit for different reasons, and also reasons that perhaps weren’t the ones the page/site creators considered when they created the site.   (It also means that this posting has the effect of causing a blip in the traffic for both sites, and their metrics person will have to figure out why this happened and will be wondering why a blog on metrics relates to butterflies…  Chaos rules!)

I point this out because it’s important and necessary to understand from the get-go that any information involving web metrics completely fraught with assumptions, and even more murky for most non-commercial sites.   The upside is that most everyone is in the same situation (or just isn’t particularly concerned).

So what can be done?

Get used to looking at where the referrers to your site are.   If you’re using something like Google Analytics, they’ll compile it for you in terms of whether the referral came from another web site, if the URL was typed in or bookmarked, or from a search engine.     Each of these tells you something.

  1. The set of web-site referrals is interesting because it means that someone actually took the time to link to you!    Checking those sites out can be a great ego-boost (or incredibly frustrating if you’re not happy with the source of the link), but it gives you insight into additional user populations you might not have considered (e.g., the people visiting butterfly sites because they’re into web metrics).
  2. Bookmarked and typed URLs are good because it means that a particular user is going out of their way to visit you.   For bookmarks, it generally means valuable repeat traffic even if it’s a single page view or very short visit!
  3. Search engine-based visits are good in terms of knowing that your site is “getting out there”, but not always indicative of a successful visit.   While some of them are successes (even the single page views), it’s not immediately knowable which ones are successes or “nope – not what I’m looking for – try the next site on the list” situations.   (One tedious experiment I’ve been wanting to try – using myself as a guinea pig – is to track my success rate in finding information from web searches: how many sites I check out before I find what I’m searching for, how long it takes from the first search to the “aha!” moment, etc.)   For most search engines, the referral URL will contain the search terms (though you have to parse that out of the URL – again online services (e.g., Google Analytics) will take care of that for you, so it’s also worthwhile to see what people were searching on that caused them to find your site.

Anyway, let’s say you’ve got 30,000 distinct visitors last month (and 42,000 visits) according to your favorite web analyzer.   Both numbers have some margin of error:  visitors are usually counted by IP number and there are several sitations were either the same person came to the site from different IPs or several people came to the site via the same IP.   If your potential site population (estimated from above) is 400,000 people (by whatever criteria you arrived at that), then you’re somewhere around 7.5% of directly reaching your target audience.   That sounds small, but consider that the mom-and-pop coffee place up the street probably only gets a small fraction of the number of coffee drinkers in the area (however you’re defining “the area”) let alone the whole planet.   PLUS, it’s a baseline that can be subsequently monitored/checked for growth.

Now there’s the other side of the equation: where did those visitors go on your site, and how do those locations match up to how they got there?   We tend to design our sites from a “top/down” perspective with the assumption that people will enter through the “home page” and then winnow their way down to the content.   The funny thing is that that assumption becomes less valid for repeat visitors – once they’ve become familiar with your site it becomes more likely that they’ll bypass any extra clicks that it takes to get to the content they want.   Similarly, search engines don’t really care about your home page if they’ve crawled your site and are able to effectively get visitors  to inside pages.

So begins the messy investigative work of trying to connect the “from”s and “to”s in a way that will draw you a picture of how things are working.   But even a shallow exploration is likely to reveal surprises.  Plus, as you become more accustomed to diving into the log files (or whatever way you’ve assembled that data), you’ll start to get a sense of what sort of “patterns” to look for.

Let’s Share! What interesting/weird/exciting traffic has your site has from a source you didn’t expect?   Has it changed your thoughts on how you’ve designed your site?


September 10, 2010 - Posted by | philosophy, visits |

1 Comment »

  1. Update: I just discovered that this post was reached in a web search for много бабочек фото (Russian for “many butterfiles photo”)… 🙂

    Comment by Bob Donahue | September 17, 2010 | Reply

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: