College Web Development

Jump to Comments

So you have a wife and two kids and you can not afford to lose your job. Good, you are in the same boat as the rest of us. However, if you are a website developer and you are not testing with an optimizing software like Google Website Optimizer you will lose your job, or at least you should. This post talks about the best ways to start testing your site to improve conversion rates, making it impossible for you to be replaced. At the end I will even touch on ways to a promotion, assuming your boss is not a jerk.

If you have ever said the following, ” My website is < insert some claim to complexity>, so there is no way I can test it.” you are wrong. Your website is not too complex to test, you can do it and you should have started months ago. There is a good article by John Quarto-Vontivadar titled The Price of Perfection, which touches on the importance of not lying to yourself and start testing.

What is website optimization?

First, if you are asking this question you are a step ahead of the game. Around 3/4 of e-comerce websites do not optimize their websites, so you are one step closer to increasing your job security. Now, take a look at the resources on the Google’s Website Optimizer website. In summery, you conduct tests to see what causes more users to convert.

Where do I start?

The starting point for college websites is unique in that most college websites lack concrete goals. Read my post on Excellent College Website Goals to learn how to establish goals for your site. For this post I will refer to Future Students section of the website for it is a site with high traffic and common goals for most colleges.

What should I test first?

There are lots of blog posts about what you should test first, here are two resources to help you get started: Prioritize Optimization, Introduction to Website Optimizer Webinar. For your first few tests, go to your future students page and find some major text or photos, then swap them with a couple other versions. Do not do too many versions at first, get a little experience under your belt before you do any complex testing.

What to test for a college website:

• Future Students landing page
• Apply Online buttons
• Request a tour button
• Test, offering incentives for tour request or information request forms
• Test the application process - this is a big one, if your application process is complex you will lose completed applications.

There is no limit to what you can test, so look at your pages and try new ideas. And remember if anyone has an idea during a meeting, do not allow that idea to be wasted. Just say you will test it to see if it increases your conversion rates.

So you don’t just want to keep your job, you want a promotion!

Now this is where optimization gets interesting, you want to increase user satisfaction so how can you do it, well a survey is the only real way to see if your users are enjoying the experience. How are you supposed to test if people are enjoying one version of the site or another?

This is more complex and less objective then normal optimization tests but in no way should this be ignored for college websites. Because every college website should have a goal of user satisfaction, often times it is difficult to find a goal page for a certain site, like the advisement office site.

How to execute it: simply create an a/b test with 2 versions of the page, and use any page as the goal page. Then include different surveys with the same exact questions, i.e. the same survey but the results go to different locations. This test will take a while to execute because you need enough data to have reliable information. Run the test until you feel there is significant data, best is 2 weeks to 6 weeks. Keep the surveys short no more than 6 questions, and one should be “What is your overall opinion of the site?” You can also use 4Q, which is a decent survey for such a test.

Optimize now and forever

Your site should constantly be tested and optimized. Your users change year after year and what might increase your conversions one year can decrease them the next. It can even be seasonal, don’t keep snowflakes on your page from a test in November when it is July.