Monday, January 21st, 2008

Three Reasons You SHOULD Use Dates In URLs

There was a post by Aaron Wall the other day on Search Engine Land entitled "Four Reasons To Avoid Using Dates in URLs". I'm a big advocate of dates in URLs in certain circumstances, so let's look at things more closely - starting with why Aaron thinks you shouldn't use URLs...

Aaron's Reason #1: "Irrelevant information"

His point is that dates are irrelevant information to the searcher and since the date often comes before a descriptive file name, you're putting irrelevant information ahead of relevant information. While I can understand that dates may be irrelevant to say a cooking site, they're definitely relevant for topics such as technology, politics, search engine optimization, community development, automobiles... The list goes on and on... How many times have you looked at an article and wished you knew when it was written? You want to know if something is obsolete or not and dates convey that information. Personally, there are a lot of times when I even want to see them in the body of the post.

But he's also saying that directories are more important than file names because they come first in the URL. If I had to guess, it's just the opposite. URLs work from the least specific item (the site name) to the most specific (the file name). In the URL for this page is not more important than the name of the file. He may be speaking of apparent importance to the user, but I find most users think of URLs as 'noise' and ignore them. The one exception is particularly long URLs - but they're a separate issue and should to be avoided generally...

Aaron's Reason #2: Dates Suggest Things are Old and Stale

That's exactly why I love dates - they tell me what's old and stale. To pretend that your content is evergreen when it's not isn't fair to the user. Sure, some of what you publish will be evergreen, but let the content speak for itself in that regard. Trying to hide the fact that the page is stale isn't fair to the user or the search engine who just wants to find the best information on a topic.

Aaron's Reason #3: Dates Make Sharing URLs Difficult

The longer the URL the more likely it will be to break when you write an e-mail. Sure... Maybe... But so do long domain names and long file names. On this blog the date portion adds just 8 characters. I thing that's a fair price to pay for a more descriptive URL. But sure - keep your URLs short, if you can... Still, even if they're long, the sharing issue isn't a huge problem.

Aaron's Reason #4: Dates Can Cause Robots.txt Errors

Guns don't kill people, people kill people... Same with dates in URLs - if you're writing a robots.txt file you better think about the consequences of what you're doing. It's something that has to be done carefully and monitored after the fact. Robots.txt is case sensitive which can cause some major problems if your server isn't case sensitive, and there are rules on which spiders need to see what that you wouldn't necessarily expect - like multimedia spiders needing to see HTML, not just multimedia files.

So those are his 4 reasons - needless to say I'm not really buying his logic. Here are my three reasons why you should use dates in your URLs:

My Reason #1: Dates Suggest Things are Old and Stale

People aren't stupid. If they're looking for a good recipe they know they don't care about the date. If they're looking at a blog from a sailing charter captain they know when the date matters ("I'll be in the Windward Islands in a month") and when it doesn't matter (searching for photos of the Windward Islands). And there are lots of cases where the date is critical (e.g. bugs in software).

Putting the user first and giving them all of the relevant information about the page is important. Treat your users right and they'll come back.

My Reason #2: A Maintainable Site Structure

There are some prominent sites that are severely flattening the structure of their site by removing all of the directories and putting all of the "files" in the root directory of their site. So, instead of they'll have something like - they eliminate all directories plus the file name extension.

I dread the day someone asks me to take over one of those sites - especially if it's been an active blog site. Flat directory structures are not maintainable long-term. Directories exist for a reason - to have a manageable number of files in each directory and to make those files easily portable. With the advent of database driven sites (e.g. blogs based on WordPress), the number of files in the directory isn't as relevant as it used to be. But there will still come a time when you'll want to do something with all those old blog posts, and if you have a flat directory structure, you're in trouble because there's no easy way to distinguish old URLs from new ones without a database lookup.

At the same time you don't need to go crazy. I've seen options in some blogging software to have directories down to the day of the month (e.g. /2008/01/21/). That's ridiculous even if you're a relatively active blogger. One directory per month is more than sufficient (e.g. 2008-01) and will avoid problems with your directory structure being too deep - which is something some search engines don't seem to like.

My Reason #3: File Name Reuse

Last year Google Analytics was redesigned. The people who have flat directory structures will have to choose a different file name next time Google Analytics is redesigned in a few years. In other words, is already taken. But they could have had and then a few years later - two "files" with the same name in different directories. And gee - the URL tells the user which one is the 2007 redesign, and which one is the 2009 redesign... Doesn't seem "irrelevant" to me.

That about sums it up... There's a lot to be said for using dates in your URLs - just be careful to use them judiciously.

Categories: Web Site Configuration

Leave a Reply