3 Reasons Why it is Important to Test Drive your Site

This is a question you should ask yourself for each site you have - how well do you know your own site?

If you’re answer is anything other than “like the palm of my hand”, you should immediately go and learn every aspect of your website.

Webmasters set up a new site, install a nice template and make some modifications to it for a more unique feel, and then start writing content. But not all of them give their site a full “test drive”, and end up missing something important - I know this because it happened to me on this very site.

Why is it important to test drive your site?

1. You need to make sure that structurally your layout displays correctly for all users with all browsers on all pages of your site. Obviously, you can only do so much of this yourself. This is where it is handy to ask fellow webmasters or friends to give your site a whirl and see what they think – and who knows, maybe they will like it enough to subscribe.
   
2. You need to see from the perspective of the user – sure, your design may make sense to you, but is it confusing to the user? The average reader won’t spend more than a couple of seconds looking for what they want, and if they fail to find it, they’re gone. Make sure your linking strategies are obvious and all of your site’s important options (like subscribing to your RSS feed) are clearly displayed to the user on all pages.
   
3. Notice what is not on your site. Are you missing a search function? Maybe your header logo does not link back to your index? Is there no way for a reader to get in contact with you?

This is where I made a mistake. On my homepage, there is a large eye-catching button for my RSS feed on the side menu bar. However, I never noticed that when viewing a single article, the button no longer displays. In fact, there was no way a user would even know that I even had a RSS feed, let alone be able to subscribe to it.

Here is the real disaster – all of the social bookmarking sites that index my articles and point interested users to them link directly to the single article, not the home page. So when any of my stories, for example, get traffic from Digg, the users are being taken directly to the article where I am not giving them an option to subscribe. Most of them are reading and leaving, whereas if I had that RSS button displayed they would have been far more likely to become repeat visitors.

Think of this: it’s real easy to click one button and subscribe to a site’s RSS feed and “preview” the content for a couple of days to see if you want to become a regular reader. But if you have to instead bookmark the site (most people already have more bookmarks than they’ll ever visit) and remember to visit it again in a day or two…well, let’s just say that the odds are not in your favor.

So, the moral of this story is to learn your site insides and out – you may be surprised with what you find. Think of all of the potential subscribers I could have missed out on if I had not caught this over site so early.

 

An interesting read. thankyou.

 

Makes me want to do some user testing. Very inspiring.

 

Testing is something a lot of people look past, and I agree with you completely. Regardless of whether you are coding or writing, you should always run several different scenarios/test cases with your site. Just because it works the way you want it to work, doesn't mean it will work/break when something out of the usual is entered

 

I agree with you in principle however User Acceptance Testing (UAT) is exceptionally difficult to do objectively if you have been in any way involved in the design/ development as you will have lived and breathed your site and so the interface will be intuitive to you and you are still testing against your own view of what is correct rather than the general population.

We have the luxury of being able to hand over UAT testing to someone who hasnt worked on the project but even then there will be certain blindness to certain things as any company will develop its own standard methodologies and so these will almost certainly just be accepted during UAT rather than being challenged.

It is generally better to get effectively random people who know nothing about your project to do this kind of testing as they will be much better at telling you if something is easy to use etc

 

I completely agree user testing is essential for any medium-large web site. I generally do browser tests myself, W3C validation check, and then release.

 

Totally agree. I have been doing user interface testing for some time with a fortune 1000 company. The issues that we have been able to resolve before going to market are incredible. There were a lot of faux pas that were avoided by diligent testing. We were also able to enhance usability of the sites we were working on. I am actually thinking of venturing out on my own as I believe that moving forward a lot of us (and we are not Fortune 1000's YET LOL) will see the importance of UAT.

 

Testing is pretty hard. After reading Steve Krug's book Don't Make Me Think, I thought okay, when I build sites I'll try to look at it with fresh eyes (this was before I did anything other than GIMP mockups). Well, I can't. I need completely unfamiliar people to try to do a particular, set TASK on my site. Where did they go first? What did they think? What text was ambiguous? Did they miss the big huge button with the name of the task written in huge flashing letters on it? (Someone on SitePoint posted a Neilson link where people were asked to find the current US population on the Census Bureau site. The current population is written in HUGE RED LETTERS on the right side of the site, but only 14% of the tested found it!)

I also listen to my sites in JAWS. However, I'm really not very good at using it. You can do all these customisations and make all these shortcuts with it-- I'll never know how a real user with JAWS will experience my site (plus I only listen to see what it says, and I haven't taken it out of the default of English, and my site's in Dutch, so it sounds pretty funny).

Testing is really important, but as AstarothSolutions said, it's hard, longduring, and expensive. Like, it can be really really expensive. If you're having your friends test your site, choose friends who only surf to check their email and look up stuff on Google-- don't use the tech-savvy unless they are your target. Give them specific tasks to do! Can grandma do This That and The Other on your site??? What if she never saw the main page? Can she still do the task?

What we can do is check our sites with multiple user agents-- screen readers, computer browsers, mobile browsers, Javascript on and off, CSS on and off, different screen resolutions, differeny colours (like Opera's High Contrast Option), and while your at it, send your site to Vischeck.com or one of those sites and see how it looks to the colourblind. Were you relying on colour to send special information?? Can your site be read when the colour contrast isn't there due to common sight problems? Does your layout totally break and look like crap when someone increases the text size once or twice? For mulitlingual sites, is it easy for someone with only one language to find that button that changes the site over to their language? Cause they can't read anything else on the site to find it.

Course, all this is fine and dandy when you aren't given a project with a too-short deadline : )

Nice post, Charlie Brown : )

 

very nice information
testing website is very important...
i always remember to do that

 

The absolute most basic tests that should be done on every form and each and every single input element:

1. Buffer Overflow - Verify you do not allow more characters then what the database has.
2. Special Characters - If POST'ed data comes back around to the form/onscreen, be sure the string/special characters are properly encoded.
3. Database inserts - verify every single variable entered into the database is properly protected against SQL Injection

These are the base three things I look for in testing before I even start checking if it is the proper data / data type is coming through.

 

For sure agree that you should test your site. That is what I do, Usability and Design Testing. I do a basic one that even gets a free follow up review. By the way, your site: http://www.abstractpromotion.com/, at first I thought it was missing an introduction message. I finally found it just under the logo, at about 4 microns tall. Seriously, you might want to make that stand out a little more. I had no idea what your site was about at first. Also, you should set your font size to em or pixels. Many people like to adjust the font size to read easier, and IE7 will not let you do that unless your text size is set to a relative setting, and not px.

 

Thanks for the suggestions, I appreciate it

 

No problem. I try to put out new tips every day on the site. Here is a link to RSS if you are interested: http://feeds.feedburner.com/BestWebImage