Most Recent | 2008 | 2007 | 2006 | 2005 | 2004 | 2003
Engineering Creative Freedom
Red Door Interactive's Kelly Abbott shares easy ways to test your site without paying a mint or waiting around for your IT department.
iMedia Connection
Kelly Abbott
When people ask me what it is I do, I tell them, "I create engineer-free zones." It's my way of saying to marketing professionals, "give me the vision. I'll try to make it work." This can be a refreshing approach, since often the creative and technical elements of marketing butt heads -- each insisting the other is preventing it from doing its job properly.
While Reid Carr is on paternity leave, I'll be hijacking his regular column to, I hope, arm you with information to test your site and get your way with the technical team. In time, you'll be creating engineer-free zones for your own internet presence.
Let's start with the following three engineer-friendly scenarios.
Scenario 1: On a Mac, your site looks like garbage. You know because your friend has one, and you've seen it with your own two eyes.
I love this problem because it's so easy to diagnose and solve. There is a beautiful service that will take screenshots of any page on your website from any OS, browser and resolution, with or without Flash. If you need a specific combination of any of the factors above, that's no problem. But more than that, the process is web-based and need not involve anyone from your technical staff to set up.
To demonstrate some examples, I set up an archive of a test I just ran on three Yahoo! domains: http://www.yahoo.com/, my.yahoo.com, and maps.yahoo.com. The monitoring service is called BrowserCam. You can take it for a free test-drive for 24 hours, and it took about five minutes for me to log in, select my URLs and get a finished report. The report is public, so you can have a look for yourself. What you'll find is a lot of screenshots of the same three sites from some of the (OS, browser, resolution) combinations above.
This capability enables you to look for inconsistencies that may be affecting the way a certain subset of users "views" your site. A common issue is how styles are displayed in Internet Explorer versus in Firefox, Netscape, Opera and Safari.
Internet Explorer on the Mac is a particularly troublesome browser when it comes to styles, so watch out for that one. You'll also want to see what happens on lower resolution computers. Many laptops are easier to read when the resolution is set low, so lots of travelers and people who work remotely tend to lose important information "below the fold." If you see critical content drop off the first screen grab at lower resolutions, you may want to make modifications to include that information higher and improve the stickiness of your homepage.
Scenario 2: Customers are getting close to converting, you think, but you can't tell if they're having technical problems or are just chickening out before hitting "submit."
There are two factors to this test. On one hand, you need to make sure that the form works; not only now, but over time. Once you can confirm that it is consistently "submit-able," as it were, you need to determine why people are abandoning the form.
First, you'll want to subscribe to a website monitoring service like Webmetrics (you can get a 30 day free trial). This robot's sole purpose is to try to complete a simple task repeatedly, 24 hours-per-day, seven days per week. It automatically submits your critical form once every twenty minutes, and will email you, call you and SMS you when it encounters an error, so that you can act on the information.
What's great about this service is that it does not require you do anything technical on your end to set it up. You don't have to install software or call your IT department to let them know you're doing it, et cetera.
Now that you're sure your forms are working correctly, you want to see where in the conversion process potential customers drop off. This will require some help from your technical staff. What you want to do is install a "user tracking" service on your site.
User tracking is a service like log file analysis that allows you to see the basic traffic patterns on your site. But more than that, it actually tracks user behavior as opposed to server behavior. That's an important distinction here. Log file analysis tools won't help you with this particular problem. The reason user tracking tools (like HBX from WebSideStory) will work better, in this situation, is that they can see which fields are being abandoned. If someone doesn't want to give you his or her phone number, you'll see it.
User tracking is a good all-around tool for quantifying the user experience. With visual overlays and funnel analysis tools offered as a part of the same service, you'll be able to see the links people are clicking on to convert, the pages that are critical for successful conversions, and what your drop-off rates are between each pebble in your clickstream.
Last, by adding real conversion values to a successful conversion, you can see your real and projected ROI based on existing site performance and anticipated improvements.
That's the vision. The technical requirements are small but not insignificant.
Your technical staff will have to install a few lines of code on each page of your website. Depending on the number of pages and the way your site is architected, the task could take as little as three minutes to complete.
In high-traffic, mission-critical internet presences, however, adding a few lines of code site-wide is a reasonable request, and you may have bigger technical fish to fry. HBX has implementation specialists and installation guides that take the headaches out of installation for even the most reluctant of engineers.
Scenario 3: Your well-placed ads may not be all that well-placed. The clickthrough's are pathetic, and you have a hunch it's not the creative, but the placement on the page.
The humanist in me says it's tragic, the technologist in me admits it's true: the internet as a medium makes almost everything measurable.
If happens on a site, it can be measured. And in some cases, if it doesn't happen, that can be measured too.
Take Eyetools.com for example. This is a service that employs people to look at any given page on your site and respond to some basic questions. It sounds a lot like a focus group, and it is. Only instead of taking days to set up, lasting hours and costing upwards of $10,000, this is a much easier solution to provide bare bones facts on whether the message you have -- on your home page, for example -- is getting through to users.
Ten people, two days, and $1000 later, you have your answer.
It's quick and dirty, but sometimes that's just what you need to get your IT department to concede that the ad space on your home page perhaps could do with improved placement. Why argue over the unknown, when even what you don't know can be measured?
Full Story >>
