Back in the day, the energy company Powergen Italia bought a .com domain, and if ever there was a case for using a hyphen in a domain name, that was it. 

Some brands have clearly drawn the short end of the straw, as far as amusing / embarrassing / horrific domain names go, but most can avoid being known for having stupid URLs. They come in all kinds of shapes and sizes and we still see plenty of lame URLs in need of some special attention.

I’m not going to get into the technical detail, largely because not all of these points are technical in nature, and Econsultancy’s how-to-guides on SEO Best Practice and Effective Web Design contain around 700 pages of practical insight on – among other things – URL strategies.

Here are 15 things to look out for when trying to discover if your site suffers from URLitis. Take that medication if it does!

Descriptive URLs FTW

It is important to include descriptive keywords in the URL – for potential visitors, and for the search engines (it helps Google make sense of your pages if you do). Anyway, why wouldn’t you do this?

Custom URLs

Some sites use a different URL than the headline, to condense the URL / minimise the amount of words used. The Guardian, for example. Similarly, I think there might be a case for writing long keyword-rich headlines to begin with, then saving your post / URL, and then removing some of those words to: a) reduce the headline length while b) retaining important keywords in the URL itself. I made this point in my 23 guidelines for web writers, as you may benefit from a small SEO uplift.

URL keyword stuffing

Some sites cram as many keywords into URLs as is humanly possible. URL keyword stuffing isn’t a great idea…

URL keyword famine

I need more than just one keyword to understand what your page might be about. Short slugs can work though, but one word (or even two) might be pushing it for article-based pages. 

Long-ass (and totally nonsensical) URLs

Some of the biggest publishers still haven’t bothered to introduce the easy tech required to rewrite URLs to reflect headlines / content (they’re nuts). A 20-character string of numbers and letters means nothing to me, Reuters

Guessable URLs

If you know that /jobs takes you to our Jobs channel then you might guess that /blog brings you here. And you’d be right. I do this sort of thing every day on all kinds of sites. Other web users do too. 

Avoid some forms of punctuation

Apostrophes, for example, can cause trouble for users that try to share them on sites like StumbleUpon.

Underscores are NOT the same as hyphens

First they weren’t. Then they were. Now they aren’t. Use hyphens as word separators, not underscores (or anything else, for that matter).

Canonical URLs

It makes sense for sites with lots of duplicate (or very similar content) to use the canonical URL tag, which is supported by Google (I think Yahoo and Bing have yet to embrace this as a common standard). Google provides more background and detail on canonical URLs while Econsultancy’s SEO Best Practice Guide contains plenty of guidance on how to set things up. 

Referrer sources and session IDs

Some URLs seem to go on forever don’t they? They can appended with ridiculous amounts of visible clutter, showing where you came from, what keywords you used, what browser you’re using, what ad campaign you clicked, which affiliate website referred you, your session ID and various other bits of useless (for the visitor, though not for the website owner) information. Does this data really need to be reflected in URL strings? All too often I’ll forward a link including a session ID which – when clicked on by a friend / follower / the recipient – defaults not to the page I was looking at but to the homepage. That feels kinda broken to me…

URLs shouldn’t be messed around with after publication

The very last thing you want to do is change your URL. Consider those newly-generated inbound links, tweets, shares, etc. Workarounds and redirects aside, don’t go changing. Redirects are not something the average e-commerce manager wants to spend much time thinking about, much less an online writer.

Unique IDs

Some sites like Google News typically require articles to have some kind of unique identifier / post ID. The advice certainly used to be to include a unique eight-digit number in your URL, though we don’t do this at Econsultancy and have no issues with indexation on Google News. But we still visibly include a unique post ID in our blog URLs. 


Dates are commonly thought to suck when placed in URLs. There are pros and cons, from a reader’s perspective, but they aren’t entirely necessary. If in doubt leave them out.


You know those hideous Flash sites that don’t assign their ‘pages’ with URLs? They are really, really lame. We live in an age of sharing, at least as far as links go. What’s a visitor to do when they want to share a deep link to a Flash page? Technically you can allow it, but technically you don’t need to build a site in Flash to make the user experience all whizzy and awesome (it is somewhat ironic that most Flash sites are truly hateful, as far as the user experience goes).

Link hovering

I commonly hover over links / anchor text to check the source of the link before clicking on it. Mainly this is because I avoid visiting sensationalist ‘news’ websites, but it also gives me a chance to see the URL in full. A strong URL can be very persuasive and can be the difference between a click and no click (good headlines / titles are compelling calls to action). 

I’m sure there are a dozen or so more URL considerations to think about when planning a new site or relaunch. What did I miss?