Why Site Tourists are Worthless
Sidebar to Jakob Nielsen's column on loyalty on the Internet, August 1997.
Tourists are users who visit your site temporarily but do not really use it. Typically, they only look at one or two pages, after which they move on to another site. This is in contrast to loyal users who visit repeatedly and may look at as many as five pages (in rare cases even more).
If you sell advertising, then you may think that higher hit-rates are a benefit in their own right: after all, more hits equal more eyeballs for the ads on your site, equals higher income. True, but only if you hold CPM constant. Any savvy advertiser will demand a lower CPM (cost per thousand viewers of the ad) if the "eyeballs" are unqualified leads who are at the site by mistake and not potential buyers of the advertised product. The only way to command high CPM is to deliver a narrowly defined set of users who are truly interested in some specific topic. Otherwise, clickthrough will drop like a brick.
The three main types of site tourists are:
Lost in the Woods: users who stumble into your site by mistake from a search engine, by following a link from another site, or by typing in an erroneous URL, or by some other means where they thought they were going to some other place than your site. Because search engines are so bad, users frequently follow links by mistake: when they see that they did not get what they wanted, they are out of the site as fast as they can get their mouse to the back button. They are not going to stay around to explore since they are on a mission to find something else. Therefore, it is self-defeating to try to lure users by misleading search terms or links that promise but do not deliver: since users control their own navigation destiny on the Web, they do not patronize sites that trick them.
Siteseers who check out your home page (and possibly a few additional pages) because they want to look at different websites and their design. Such users are valuable if you are a Web design shop, but if you sell widgets, you won't get much business from people who are more interested in your page layout than in widgets.
One-timers who go to your site to see one page that interests them but don't look around for more. Such users may recommend your site to others and they may also leave with a reasonably favorable impression of your company. Even so, if they don't return, then they obviously didn't feel that good about you. It is not likely that a one-timer becomes enough of an advocate for your site to actively drives more traffic to you; it is also not likely that a one-timer will ever buy anything from you.
People who click on an online advertisement but back out immediately after seeing a page or two might belong to any of these three categories, depending on the specifics. If the ad "cheated" by promising something that was not delivered, then the user followed the link by mistake and will be so angry that he or she will never return. If the ad made the user interested in seeing your site design or the one page with the special offer, then at least you generated a possible lead, though you almost certainly won't gain any real benefits unless the user feels sufficiently rewarded to move on to additional pages and/or subsequent visits.
The traditional arguments from non-Internet marketing continue to hold: it is almost always much cheaper to retain satisfied customers and turn them into repeat business than it is to attract a new, one-time customer. The first time you do business with somebody, the net result may well be a loss because of the huge cost of attracting new customers and prying them away from the competition. Also, the first order is likely small because the customer wants to test the waters before committing to large deals on an ongoing basis. Keeping the customer in the fold is cheaper than advertising for new customers, and repeat orders are likely be progressively larger. Combine cheaper costs and larger orders and you find that repeat customers generate most of your profits. This has been proven again and again in the "real world" and is almost certainly true on the Internet as well.
(See also: Reduce Bounce Rates: Fight for the Second Click .)