Commitment and Site Optimization

Reading a recent blog post from Jeffery Eisenberg (Realistic Expectations For Conversion Rate Optimization) made me once again think about how a lot of companies fail to really commit to testing and site optimization once they purchase a tool (Test&Target, SiteSpect, Optimpost, Goolgle Web Site Optimizer, etc.). Right now, I see site optimization where I saw Web analytics about 5 years ago in terms of tools and commitment.

A few years ago, businesses were ready to go out and buy the biggest and best Web analytics solution out there, without having any kind of dedicated resources to leverage the information or to ensure that any kind of best practices were being followed or developed. Now, many companies have dedicated Web analysts that can implement analytics solutions and help their businesses leverage the information contained within. Site optimization is, as I see it, about to explode (more than it already has) because companies appear ready to commit resources to the effort as opposed to just buying a solution and running with it.

Most companies do not dedicate any resources to actually making their existing Web sites better.

Most design and development efforts are concerned with developing new features or content. Instead, companies need to remember that they have a ton of content out there that could probably be performing better than it already is. After all, how often do any of us get something perfect on the first try (or the second for that matter)?

There are several things that a company can do to ensure that they are committed to optimizing their Web site:

  • Dedicated some of the time of your design and development teams to optimization.
  • Commit to designing at least 2 versions of everything that goes out. Make optimization a part of the design process (within reason of course). This is often a big challenge as designers see testing as just doubling their work.
  • Find a way to get everyone invested/interested. A lot of companies make the testing process an internal contest of sorts where everyone watches results in real time.
  • Pay your employees for coming up with ideas that improve conversion rates. After all, shouldn’t you be paying your employees to impact the bottom line anyway? Here, it’s measurable!
  • Realize that optimization and testing is just as important as your paid search and e-mail marketing efforts. All require an ongoing commitment in resources and effort.
Do you have any other thoughts on what companies can do to ensure that they are committed to site optimization and testing?

More on Pages Not Being Worth Anything


Followup to “Pages Aren’t Worth Anything” from Jason Egan on Vimeo.

This is my first attempt at a video post here, so this might be a little rough. They should get better though!

Summary:

  • Pages aren’t worth anything, but what you’re selling is.
  • Testing and optimization should replace analytics in determining how effective a page is at conversion.
  • You should me optimizing and measuring the worth of your marketing accumen as opposed to some HTML. Look at a page or site this way. It is a marketing and sales tool. How good is it at marketing and selling? How good are you and making the site better at marketing and selling.

Your Pages Aren’t Worth Anything

I’m sure that some of you will disagree with the title of this post as soon as you see it, but hear me out (read me out?). Wether you are in publishing/content or straight up e-commerce, your pages are not making money inherently. Many executives and busines owners always have one of the two following question:

  1. “How much revenue has this page made?”
  2. “How well is this page converting?”

To preface the rest of this post and explain the title of this post, I am going to say that pages do not make money, and they do not convert. Crazy, I know. But, what does make you money and convert are the changes that you make to your pages.

Answering, “How much revenue has this page made?

 

I do not want to turn this into a debate about revenue attribution at the page view level, so I will leave that issue for another time. So let’s just assume that you have some kind of repor that has page names and dollar amounts next to those pages. The answer to the above question is:

“Does it matter what that revenue number is if you aren’t changing anything?”

So you can trend a page’s revenue over time. So what? Let’s say that you sell hair dryers (why was that the first thing that popped into my mind?). Your sales of hair dryers are what make you money, not the view of some page on your site. Afterall, you are selling hair dryers, not page views.

I also want to take this opportunity to address the publisher/content sites out there. News flash, you are selling something! Your selling ad views, and video ad plays, not page views or time spent (or “engagement” of all silly things).

You should not be asking how much a page makes for you. You need to be coming up with ideas that you think can make a page better, and testing those ideas to see if you can create lift! A page that just sits there and is never, or blindely changed isn’t doing you any good. Does it really matter how much you think a page makes over time if you’re not trying to make it better to begin with?

You changes, improvements and efforts make money and create lift. A page sitting there isn’t doing you any good.

Answering, “How well is this page converting?

 

Again, I would say that a page just sitting there is never converting any different that it ever has, so tracking the conversion rate of a page is pointless. You should be tracking how good YOU are at making changes that improve conversion.

Lift is as Important as Revenue and Conversion

 

Just like I feel that “engagement” is an excuse on the part of publisher/content sites, I feel that tracking how much a page makes or how well it converts is an excuse for not testing your pages and working on creating lift. If you really care how much a page is making or how well it’s converting, then you should have a hypothesis as to how you can make it better, and you should test that hypothesis to create lift.