You've worked nights, alienated your loved ones, and violated every one of your New Year's resolutions to build that sparkling new Web site. Or maybe you banged out that new site using all the know-how you gleaned from Webmonkey tutorials, which has allowed you to sleep well, nurture your relationships, and make it to the gym three times a week.

Whether or not you're ready for the beach this summer, you're certain that your Web site is completely ready for prime time. But is it? Who besides you and the toys sitting atop your monitor have even seen the final product? Maybe you've shown it to a few of your closest friends. Or perhaps you've asked a couple of people to navigate your pages and look for JavaScript errors. Perhaps you've even sent the URL to your mother, who couldn't quite manage to load the frontdoor using AOL but says "it looks wonderful."

Any input is certainly good input, because we as Web developers tend to lose our perspective on projects we've invested so much time in. Any Web site you create is imbued with your assumptions and affinities about design, interface, and product requirements. After spending a month clicking that oh-so-clever, frosted-chocolate-doughnut help button or letting your brain adapt to the florescent-green background, it's impossible for you to take a step back and critically evaluate how these aspects of your Web site look to a first-time user.

The friends-and-family approach is certainly a good way to get friends and family in on an IPO, but for usability testing, it is far better to watch a few members of your target audience click their way around the site. For that final look-see, there are a couple of choices: You can pay a professional company a bundle of cash to conduct user testing, or you can do it yourself. By this point, you must know that we're all in favor of do-it-yourself qualitative research. We especially like usability testing, in which developers have the opportunity to sit down with individual users as they surf the site.

Unlike focus groups, the very nature of which bulges with people, user testing of Web sites works best one on one. Similar to focus groups, pizza and beer still apply, as does the method of recruiting participants. If standing on a street corner handing out fliers isn't your style, there are a few other effective methods for recruiting participants. Use your site's existing email lists or login data to find out which users live in your local area. Contact these users via email (assuming this is within the privacy agreement you presented when they initially gave you their addresses). You might be surprised at how many eagerly agree. If you don't have existing site data, put a link for "Free pizza and beer!" up on your site. The linking page, which will solicit participation in your new product test, will probably be your most heavily trafficked page.

Here are a few helpful hints for conducting your own usability testing:

1. Don't lead your tester through the site. Usability testing is about evaluating your site's ability to speak for itself, conveying the critical issues of navigation and product benefits to the user. As much as it pains you to watch users struggle, try to avoid grabbing the mouse and yelling "Click on the frosted chocolate doughnut, you moron!" Remember, the customer is always right. If they can't find what they need, perhaps you should consider making adjustments. You can, however, ask probing (not leading) questions to gauge users' expectations of navigation elements or product features before they click on them. But ultimately, the site needs to speak for itself; you won't be there to help each new user get settled in. And if you see a user passing through pages without noticing critical features or navigation tools, these features might need revamping.

2. Keep the sessions moving. Now we're going to contradict ourselves. Sometimes, for the purposes of making the best use of the time, a little coaching may be necessary. If, through earlier sessions, you believe you've identified a major design issue that's troubling almost every user (but you haven't been able to fix it between testers!), assume that you'll change the problem spot; therefore, move users past it if they founder. Otherwise, you'll have spent two nights and $200 in beer and pizza to discover nothing more than your frosted-chocolate-doughnut navigation element wasn't nearly as clever as you thought. As far as the rest of the site goes, well, no one managed to get to it. When things like this come up, make a note, and then keep things moving.

3. If possible, let natural usage patterns occur. If you have a decent number of participants, a great thing to do is observe the different usage styles of your testers and see how your site stands up against them. Some users might look for immediate access to interesting content via a "search our site" feature. Others might use your main menu. And still others might look for the site map. Remember that all kinds of users are out there. It's a good idea to see how each type is (or isn't) satisfied by your creation.

If you don't have the time to conduct user testing yourself and have some extra money to spend, here are some companies that can help. Still, since you're the one who will be implementing any changes, we think it makes a lot of sense to be involved in the process and hear users comments first-hand.