Good URL structure

Written by Nexcerpt on May 18th, 2010 in Patterns & News.
Tags: , , , ,

This is the first of a series of posts about good URL structure.┬áMany sites have very poor URL structure — not necessarily bad content, or broken pages — URL structure that sucks. Some of it sucks enormously. Before citing examples of what’s wrong, let’s consider the benefits of a clear, organized URL among your readers and prospects online.

Frequently, for some thoughtful person who does not know you or your business, their first impression of you is a URL you create. If that person wants to use their time effectively, and focus on the best thinking and writing available, they may be unable to resist (or ignore) that first impression.

Whenever I consider clicking on a link, I float the cursor over the link, and briefly consider at least three things it reveals:

  • Is the domain familiar?
  • Is there any obvious risk?
  • Is the site organized?

If I recognize the domain as one I know or trust, I’m likely to click over. (Now and then, I recognize the domain as one that no one should trust — for example, an alternative health or political site that serves as a front for wacky conspiracy theorists — so why bother going there, other than for a laugh?)

If the URL is clear and unambiguous, I’m likely to click over. (I tend not to click on shortened URL’s — for example, from or tinyurl — unless I know and trust the source referring me there. This is partly a security concern, since they could redirect my browser anywhere.)

If the URL reveals structure and transparency — such as a recent or specific date, and a readable string — I’m likely to click over. (I may see a hundred links in a few minutes of browsing. I may consider clicking a dozen of them. When a URL communicates information to me, it wins my time and attention.)

Very briefly, float your cursor over the link recommended earlier, about URL quality. Realize all you can discover about whether you would visit, trust, or spend time on that link.

  • It’s from Google; you recognize the source.
  • It’s in /support/, so intended to be helpful.
  • It’s for /webmasters/, so possibly technical.
  • It’s supposed to be an answer of some kind.

My three questions are answered: I know the domain; there’s little risk; and someone has organized the data. Even if your questions differ, you also have some sense for whether you’d want to visit that page, or not.

I also need to offer an example of what can go wrong. Let’s pick on someone else big enough not to be offended!

Let’s say you saw reference to Robert Half — quite a substantial firm — and very valuable work they’ve done on Salary Studies in the US and Canada. Float your cursor there, and get a load of what they want you to click:

It won’t even fit on one line! In this case, the explanation doesn’t seem to be a lack of technical awareness, or attention to server configuration. Something is occurring (they use Apache; perhaps some flavor of mod_rewrite) because you actually arrive at yet another, more horrifying URL:

That URL doesn’t even reflect the cookie values they set in my browser for both JSESSIONID and UnicalID. Now, with all due respect to what must be a group of highly competent business professionals, ask yourself a few questions, based only on what the URL undeniably reveals.

  • Do you feel confident that Robert Half is well organized?
  • Do you sense that Robert Half is technically competent?
  • Do you believe that Robert Half is a good source of data?
  • Do you know that Robert Half wants to be clear and helpful?

I do not. Most everything else about the site, and its contents, suggests that I should feel confident. It appears that they’ve spent considerable time and money intent on instilling such confidence.

However, the URL indicates otherwise, and powerfully so. It tells me unpleasant things about Robert Half:

  • They may not think systematically.
  • They do not fully understand technology.
  • They are not very disciplined with data.
  • They are not concerned about my experience.

I would not click on that link. I mean that: if I really wanted the data, I would much prefer to visit their home page, and search for it. Given a choice, I simply would not click on a URL string with such bizarre content.

Does every URL you publish make a good impression — or at least not make a bad one?

In subsequent posts, I’ll give more examples of what can work against you, and suggest ways to make it right.

Comments are closed.

Recent Posts and Other Categories