What do you do if there appears a ’404 Not Found’ error message on a website? Do you close the window? Are you going to search for the right page? Do you contact the webmaster? Most people do leave the page. Their first visit to the website was also their last.
In this article I will explain you as a webmaster or marketeer how you could prevent these messages based on a simple road map. I will also give some tips on how to turn a disadvantage in to an advantage: a 404-page can be successfully deployed in the sales of your website!

The full overview in 7 steps:

1. Optimize the existing error pages
2. Improve internal links
3. Improve external links
4. NEVER remove pages from your site, but archive them
5. Redirect some common misspellings
6. Monitor error pages
7. Monitor search queries

Plus two bonus tips:

1. commercialization of the 404-page
2. keep your content up-to-date

The road map, Part 1: optimize existing error pages

Step 1: optimize existing error pages
On the web many articles can be found which explain how an error page can be made as user friendly as possible to decrease the percentage of visitors who leave your site without viewing more pages. The general view is that a 404-page must be user-friendly and that you should help visitors as good as possible by finding what they sought for. Place a few popular links or a search field on the error page. The other recommendations have already been extensively discussed at various sites, so I give merely a few links:

Additional information:

• General information can be found on Wikipedia and Evolt.org. More technical information can be found on the web designers magazine A List Apart.

Step 2: improve internal links
Prevention is better than cure. Try to avoid visitors to stream at the 404-page as much as possible, how good this page may be. A comprehensive scan shows if all links within your site (both internal and ‘outgoing’ links) are still functioning properly. The free program Xenu Link Sleuth is an excellent tool, but there are obviously more tools that do the same. The inventory of your internal links can be done automatically, but the corrections are mostly done by hand.

Additional information:

• The tool Xenu Link Sleuth. A simple Google query provides enough other tools;

Step 2.1: adding favicon.ico
This part actually belongs in step 2, but I thought it was important enough to mention it separately. Internet Explorer asks for the so-called ‘favicon.ico’ file at each Web page with a browser view. This is the small icon on the left side of the address bar of a Web browser, you must have seen it sometimes. If there is no file called ‘favicon.ico’ in the root of your website (http://www.yourwebsite.com/favicon.ico) it won’t be found and Internet Explorer generates an error message. You can easily prevented this common error by placing a favicon on your site.

Additional information:

• Technical and substantive “how-to” for dealing with favicons: http://www.trilithium.com/johan/2005/02/no-favicon/

Step 3: improving external / incoming links
Once if the internal link structure is okay, the next step is to improve the links from other sites to your site. These are the so-called inbound links or links. The inventory could also be automated, for example, with the free program back link analyzer, with the Yahoo Site Explorer and Google Webmaster Tools. The correction of incoming links is possibly more labor intensive, because you are relying on third parties. However it is worthwhile to take action on this part.
Tip: If you’re improving your external links, make sure the incoming links contain decent descriptive texts with keywords (for example: “web design and seo weblog” instead of “weblog” as link text) .

Additional information:

1. An excellent tool unfortunately with limited export opportunities is Back link analyzer;
2. Yahoo! Site Explorer: gives accurate information about incoming links;
3. Google Webmaster Tools gives an overview of all incoming links for a website. These data are not available for third parties, so you should set the webmaster tools environment with a Google account (is a winner anyway).

Step 4: NEVER remove pages, but archive them
On the news section of your site is an article from 2006. When it was placed, there was pretty much coverage by the media and the article was linked by several blogs that discussed the content in detail. But anno 2008 the news is not relevant anymore. Whatever you do, do not delete the article! The article is a while on your site already, it is included in the index of Google and Yahoo and has perhaps built up some pagerank/relevance/history. It would be an eternal shame to throw that away. So, keep it! Put it in an archive; if necessary, put an additional comment on this article that the news might be outdated and comes from the archive. This also prevents errors when people have followed the old link.

Additional information:

• Cool URIs don’t change: http://www.w3.org/Provider/Style/URI

Part 2: installation of extra usability

Step 5: redirect some common misspellings
Are certain misspellings often made on your website? Is the word ‘contat’ tapped in by many visitors when the search for your ‘contact’ page? This is a random example, but you know what I mention. If this occurs frequently, you can redirect visitors automatically (with a so-called ’301 redirect ‘) to the proper ‘contact’ page. You can find a clear overview of the different 301-techniques on the Onetomarket site.
Please note: build in an additional warning for those who have landed on the page by using a ‘misspelling’. Show a message like: ‘We have automatically redirected you to this page, while you searched for the page ‘mispelings.html’. Make sure if you want to bookmark or link this page, that you use the correct address. Or something similar to this effect.

Part 3. Aftercare: monitoring

Step 6: Monitor error pages

To have access to the amount of visitors which comes into contact with error messages on your site, you must use your statistics package.
With the free Google Analytics package, you must set a little by your self. For example, are certain misspellings made (see step 5) and which websites provide false links that causes’ bounces’. This information could also be provided through Google webmaster tools, which shows so called “crawl errors’. WordPress users can use log-404. This WordPress plugin logs error messages and makes a network overview of the last 100.

Additional information:

• Tracking 404 pages in Google Analytics;
• Debug using Google Webmaster Tools;
• And a WordPress plugin that effectively logs error pages.

Step 7: Monitor search queries
Another aspect that can be monitored, concerns the so-called referrers. Most statistical packages would not only provide information about which site provides you visitors, but also, as they came through Google, which keywords they had used in Google when they found your site. Google Analytics works in this context with ‘keyword considerations (‘All reports > ‘Marketing Optimization’> ‘Keyword considerations ‘).
These ‘considerations’ are nothing more than suggestions for keywords that you could use in your site. People were looking for these and thought that they would find them on your site. Why not give them what they want? At least, as long as it’s relevant.
We are not very satisfied with the operation of this feature of Analytics, but because anyone can use Analytics for free, we mentioned it here. Other, but from better quality and also more specialised are the following tools: HitTail, 103Bees and the Search Meter WordPress plugin.

Additional information:

• Wikipedia on referers;
• Search Meter WordPress Plugin. Keyword suggestions from the WordPress interface;
• HitTail: http://www.hittail.com/
• 103Bees: http://www.103bees.com/

Part 4. Bonus tip: 404-pages as part of the sales process

1: commercialization of the 404-page
In some cases, an error page can be used to encourage the sales on the site rather than merely provide navigation to continue to search the site.

A good example is the 404-page of Apple. At first glance, there is a ‘normal’ 404-page, there is a slightly accusing text which says that I as a visitor might have typed in something wrong (there is some improvement needed) and there are links that leads a visitor deeper in to the site.

The most striking part of that page is a trio of images of approximately 175 by 150 pixels with information about certain offers, promotions or new products. So called banners, and they rotate. For example press several times on F5 (Ctrl-R for Safari) and you will see that each time a different banner will be shown on the page. That is a very clever action from Apple. On the 404-page are teasers that are popular. Apple makes a virtue of necessity and this principle can be applied to all kinds of sites!

Additional information:

• Apple error page: http://www.apple.com/error.html

Tip 2: keep the content up-to-date
Suppose your company site is growing and there will be new jobs at your site. For some functions you are looking for new people continuously, while others already have been filled in. It is also possible that certain functions are fulfilled now, but they will be brought back on the site in the future.
In all cases, you are better off not to remove the vacancy pages, even though the functions are already fulfilled.
In the first place the pages build up a history in search engines, in links from job sites, perhaps even on the personal weblog of someone who has responded.
Secondly, it is a cold trick to add an additional piece of text on the vacancy page in question: “at this moment this vacancy is fulfilled, but there is always place for enthusiastic, talented hard workers” or something similar. That is a lot better than ’404 Not found’….