ellipsis flag icon-blogicon-check icon-comments icon-email icon-error icon-facebook icon-follow-comment icon-googleicon-hamburger icon-imedia-blog icon-imediaicon-instagramicon-left-arrow icon-linked-in icon-linked icon-linkedin icon-multi-page-view icon-person icon-print icon-right-arrow icon-save icon-searchicon-share-arrow icon-single-page-view icon-tag icon-twitter icon-unfollow icon-upload icon-valid icon-video-play icon-views icon-website icon-youtubelogo-imedia-white logo-imedia logo-mediaWhite review-star thumbs_down thumbs_up

Defining Unique Visitors

As our industry matures more people become involved. Unfortunately, as people enter the industry for the first time they bring some misconceptions with them. Much of the feedback I've received in recent months shows a growing trend of people misunderstanding unique visitor metrics.

This article seeks to address that issue.

Definition of a unique visitor

The first and most important thing to understand is that when we identify a unique visitor we do not know who that person is.

Knowing who someone is -- knowing their name -- is called "personally identifiable information" or PII. In other words, we have identified the person. This is not what is meant in web metrics by a "unique visitor." 

The official measure of a unique visitor is nothing more than a combination of IP address and User Agent (browser + operating system). In other words, a unique visitor is " + IE 6.01 on WinXP," not Bob Jones at 223 Sunset Drive. In actual fact, there is no official measure for a metric which identifies Bob Jones at all. The nature of internet technology makes identification of named people on the web impossible unless a person voluntarily tells you who they are. Identification of people is so rare there is no technology to report it, and no metric to describe it.

We're not spying

Identification of unique visitors is necessarily vague. Everyone at the same corporation, using the same software, will have the same IP address and User Agent, and will thus look like the same unique visitor. On the other hand, most home ISPs will assign a different IP address every time you connect to the web, so you'll look like a different unique visitor every day. For most web metrics purposes that's OK -- the numbers provide enough accuracy for commercial purposes.

The key point is that identification of unique visitors is not spying on people. We're not trying to follow individuals as they surf the web -- we're trying to determine whether our site was read once by two people or twice by one person. When we "identify" unique visitors, we are not finding out their names, we are merely counting how many people there were. It's exactly the same as a counter on a turnstile -- we know how many individual people went through the ticket gate, but not who they were. And though we'd never say it, we really don't care who they were. 

The nature of business on the web is different from traditional forms of business. Despite the talk of personalization, the web is not a pin-point service; in reality it is the ultimate mass media. Most online sites count themselves lucky if they can get one sale for every 30 or 40 visits.

To make money online you need to be measuring your visitors in the tens of thousands, hundreds of thousands, or even millions. If you have 10,000 visitors to your retail store every month, you're probably going to make good money. If you have 10,000 visitors to your website, you'll never earn a living from it.

The internet is about making a small margin on a very large volume. Expedia saw 16.2 million unique visitors in April 2005 (source: Nielsen//Netratings). If 16 million people are going through your site, it's not possible to adapt your pitch to individual people. You're measuring your audience behavior in the hundreds of thousands. Expedia's marketers don't care what Bob Jones did on their site; they can't.  Bob Jones is just one drop in a river of a hundred thousand people like him. They're watching which way the river flows, not how one drop moves within it.

Why count unique visitors?

So why do we need to know how many unique visitors we have? Why does it matter if a page was read once by two people or twice by one? It matters because this tells us how effective our online activities are. If someone reads the same page over and over again, it tells us something -- most likely that the copy is confusing.

From a sales analysis perspective, the most important reason to identify unique visitors is that this underpins measurement of repeat visitors. A repeat visitor is a unique visitor who has been to the site more than once.

Why measure repeat visitors?

Understanding your repeat visitors is where the money is. The reason is simple: it takes an average of 2.4 visits to your site before someone will buy. In other words, most people buy on the second or third visit to your site. No one buys the first time they visit. Repeat visitors are the ones who matter. Repeat visitors buy the products, and therefore repeat visitors pay your bills.

Estimates vary, but it is widely accepted that between 70 percent and 80 percent of all visits to ecommerce sites begin with a search in a search engine. When someone searches for a product they will get a list of 10 to 20 possible providers. This means that before they've seen your site, they know there are other people who could offer the same thing. Even if they find what they are looking for, even if they love your price, they will inevitably want to compare it with what's offered on another site. You can see this clearly in their click-stream data (the list of pages they read).  People will enter from a search engine, visit a series of pages, then rapidly hit the back button until they're back on the search engine. Often, if the product is a low value item, they may return 10 to 15 minutes later to repeat the path quickly until they get to the purchase page and buy.

Critical factors to understand about repeat visitors are how many visits it will take to get the sale, and what the frequency of visits is. Do people return after a few minutes, a few days or a few weeks? Generally you will find that the more expensive the product, or the more commitment required, the longer the cycle, or the more visits required to make the sale. If you're selling mortgages online, most people will compare your site with many others, and make multiple visits over several weeks before committing. If you're selling DVDs, they'll probably return to buy within a few minutes.

Improving sales thus becomes a matter of supporting your visitors throughout a sales process that involves multiple visits. The general "about us" information that orients a new visitor to your site on their first visit is just getting in the way on their second or third visit. Your site needs to have features that support people in different ways. If your system has a shopping cart, does it remember people so that it contains what they put in it on their last visit? Many people will put items in a cart during a visit just so they can see total costs (including shipping). They'll then go away and think about it.  Are their short-cuts in navigation so people who are trying to return to a particular page can get there quickly?

If sales come from repeat visitors, then it follows that when someone arrives at your site for the first time the most important thing is to convert them into a repeat visitor. Don't waste effort trying to get a sale out of them, you've very little chance. You'll get a better return on your effort by making sure the site motivates them to return, and makes it as easy as possible to purchase on a repeat visit. Ask yourself what features your site has that are specifically designed to cater to repeat visitors. Those are the features which make you money. Ask yourself what metrics you have about your repeat visitors. Those are the metrics that provide the intelligence to earn that money.

This is what counting unique visitors is all about.

Brandt Dainow is CEO of Think Metrics, creator of the InSite Web reporting system. Read full bio.

Brandt is an independent web analyst, researcher and academic.  As a web analyst, he specialises in building bespoke (or customised) web analytic reporting systems.  This can range from building a customised report format to creating an...

View full biography


to leave comments.