These Test Results Are Dense!

Something’s been going on. Some­thing behind the scenes that you may not know about. Some­thing that’s been help­ing me increase the effec­tive­ness of my web­sites and blogs.

Right under your nose. Er, I mean finger.

As you know, I’m a fanat­i­cal tester. And I test con­stantly. But what you may not know is that, it is not only lim­ited to split-​​testing copy.

Well, let me let you in on a secret.

Right now, as you’re read­ing this very post, I’m watch­ing you. Yup, I see exactly what you see, where your mouse goes and what catches your atten­tion most. (And what you read, too.)

Now, don’t worry. I’m not invad­ing your pri­vacy by infil­trat­ing your com­puter with spy­ware. And I cer­tainly don’t know who you are — nor do I care (well, I do, but not in this context).

There are var­i­ous pieces of soft­ware run­ning in the back­ground, on this web page, that track your move­ments. Not for the sake of dis­cov­er­ing who you are or what you do, but of learn­ing what appeals to you the most on the same page, what you click on the most, what cap­tures your atten­tion best, and where you are most “active.”

Let me share a few of them so you can start imple­ment­ing these tools on your site. (Using these tools, I have already dou­bled my effec­tive­ness of my site, whether it’s my ads’ click­through rates, my con­ver­sion rates or my “stick­i­ness” factor.)

One of them is CrazyEgg​.com.

CrazyEgg is a click track­ing tool. But its power is not in its data but in its deliv­ery. Surely you can track your clicks on your web­site with most soft­ware that can spit out the data (often in tab­u­lar for­mat), and ana­lyze it for you — if it’s intel­li­gent enough.

What about den­sity of clicks? What about mouse move­ments? What about attrac­tion and eye grav­ity? Heck, what about clicks on a web­page on objects that are not hyper­linked? I mean, anywhere?

Yes, you can dis­cover exactly what peo­ple are look­ing at and click­ing on, whether it’s a link or not.

CrazyEgg tracks the mouse activ­ity on your web page. But that is not its main strength. Its core ben­e­fit is that it visu­ally demon­strates, through the use of “heatmaps,” the activ­ity, pat­terns and den­sity of your read­ers’ mouse actions.

Why is this important?

Aside from know­ing which ads get clicked on, which links get more atten­tion than oth­ers, and where peo­ple grav­i­tate on the page, you can instantly see, in a snap, what dri­ves the most atten­tion — and action.

In fact, one of the most astound­ing dis­cov­er­ies I’ve made is that graph­ics (not large graph­ics but icons, pic­tures, avatars, audios and videos) tend to attract click activ­ity, whether or not they are active links.

For some rea­son, peo­ple tend to click on objects and graph­ics — they are mostly smaller boxes or pic­tures within the mid­dle of your con­tent — either to see where it leads (out of curios­ity), or to specif­i­cally do some­thing or go some­where, likely to retrieve addi­tional con­tent related to what the graphic com­mu­ni­cates or implies.

That’s not all. Not only can you track what gets clicked on the most (and that can be any area of your web page), you can use track­ing tools like CrazyEgg to test dif­fer­ent designs, lay­outs, ad place­ments (such as with your AdSense ads), col­ors, graph­ics, nav­i­ga­tion, forms, copy ele­ments, order pages, etc.

Here’s an example.

Here’s a look at what CrazyEgg tracked on my blog’s front page. The small cir­cu­lar graph­ics show click activ­ity. Notice the dif­fer­ence in color. Blue are less active ones, and the color ranges all the way to green, yel­low, organge and then red, which rep­re­sents the “hottest spots” that gen­er­ate the most clicks.

You can click on any of these objects, and it will open up a bar graph rep­re­sent­ing not only how many clicks it received but it’s ratio in com­par­i­son to other clicks tracked on the page.

This tells you what peo­ple clicked on the most, in rela­tion to all the other links — and even clicks on objects or non-​​links — on the page.

You can open these graphs indi­vid­u­ally, but here’s what it looks like when you “open all” (there’s a tog­gle but­ton to open or col­lapse all the bar graphs on the page).

Look at how it high­lights some of the most active areas.

Sweet, eh?

Of course, if you’re the more tra­di­tional “just-​​give-​​me-​​the-​​numbers” ana­lyst (or if you sim­ply pre­fer a tab­u­lar for­mat for fur­ther analy­sis), you can also retrieve a full list of all the clicks on the page, as well as their click den­sity ratio.

Here’s the kicker.

When you get CrazyEgg to trans­form the bar graphs into an over­all heatmap, that is a visual rep­re­sen­ta­tion of the “hotspots” on the page, you get a stun­ning visual overview of mouse activ­ity at a glance.

On the front page of my blog, post excerpts are dis­played with sev­eral links that lead to the entire post. One is in the title itself, another is the “read more” link at the end of the excerpt, another is the “com­ments” link, and then there’s the icon itself, which I try to hyper­link as much a pos­si­ble to the rest of the post — or some­thing related.

The dis­cov­ery? Notice that the hottest spot is the “read more” link after the excerpt of the post.

Peo­ple click more on this link to con­tinue read­ing than they do any other link. (And that’s pretty much the same for every new post I publish.)

But it doesn’t stop there.

I’ve tested dif­fer­ent word­ings, such as “read more,” “read the rest,” “click here to read,” “con­tinue read­ing,” and just plain “more.” (And a few others.)

The win­ner is the one you see on my blog right now. (But I’d also like to test arrows, ani­mated arrows, dif­fer­ent col­ors of arrows, etc.)

And nope, that’s not all, either.

I’ve tested posi­tion­ing: the “read more” link on the left, on the right, imme­di­ately after the excerpt (inline with the para­graph and not as a sep­a­rate para­graph), no link at all (the title and “com­ments” link offer the same func­tion any­way), and more.

Well, the left posi­tion, below the excerpt, won hands down.

Here’s the inter­est­ing bit: I include icons (some peo­ple call them “avatars” or post images) with each post, mostly aligned to the left at the top of the post.

As I stated ear­lier, when I started track­ing non-​​link mouse activ­ity, I dis­cov­ered that peo­ple were click­ing on these graphics.

So I thought, “Why not con­vert them into links?” Some (in fact, most of them) have the same func­tion as the “read more” link. That is, if peo­ple click on these icons, they lead the reader to the full post.

And it works!

Even when there’s no indi­ca­tion what­so­ever that the icons are hyper­linked. Peo­ple are click­ing on the graphic to read the rest of the post just as they would the title or the “more” link.

What I found most inter­est­ing, though, is this: if the graphic has what seems to be dis­cern­able clickspots, such as con­trols, slid­ers, but­tons or links (like with the tra­di­tional blue-​​underlined words), peo­ple will click them. And they will click the graphic more because of them.

Here’s a forinstance:

In one post, I talked about my wife’s lat­est video, where she shaved her head for a breast can­cer char­ity. The post icon used was basi­cally a tiny screen­shot of the video indi­cat­ing the topic of the post. That’s it, and noth­ing more. (And of course, the graphic is hyperlinked.)

But what CrazyEgg has found is that peo­ple not only click on the graphic but also on the tiny video con­trols, par­tic­u­larly the “play” but­ton — even though they don’t work! It’s as if they were try­ing to view the video right there.

What this tells you is, graph­ics with dis­cern­able clickspots are not only induc­ing clicks, but they are also appeal­ing to a more dynamic, one-​​page process.

Here’s my think­ing: if the video can run right there, on the same page, with­out load­ing in a sep­a­rate page or win­dow, then peo­ple would want to view it right there on the spot.

One of the more com­mon tools in enabling inter­ac­tiv­ity with vis­i­tors (espe­cially in the new Web 2.0, social-​​based envi­ron­ment in which we now find our­selves), is AJAX.

Basi­cally, AJAX (and its couter­part Ruby on Rails) are plat­forms and tools that allow pages to load con­tent dynam­i­cally within the page itself, with­out the need to reload.

Con­tent can fly in and out, fold up and down, and be dragged and dropped. (You see this more com­monly with a lot of the start page web­sites, such as Page​Flakes​.com and NetVibes​.com.)

Rather than sta­tic pages load­ing with each click, peo­ple would pre­fer to stay put and have the page change dynam­i­cally for them with­out chang­ing the lay­out, struc­ture or for­mat­ting too much.

Why? It’s eas­ier to inter­act with the site, it’s less dis­tract­ing and there are less changes occur­ring, and it’s seem­ingly more responsive.

(Ah yes, Web 2.0.)

Now, this is just a semi-​​educated guess, of course. But my think­ing is not just based on this one sin­gle test. Many other dif­fer­ent types of tests come to a sim­i­lar, log­i­cal conclusion.

For exam­ple, Flint McGlaughlin’s Mar​keting​Ex​per​i​ments​.com, which con­ducts, well, mar­ket­ing exper­i­ments, have con­cluded that peo­ple pre­fer more copy (and con­tent) on less pages, than less con­tent that’s spread out over mul­ti­ple pages.

Under­standly, his tests are lim­ited. While the results are widely applic­a­ble, they are cer­tainly not uni­ver­sal. But it does tell you some­thing nonethe­less: it tells you what peo­ple pre­fer, and what trends are going on right now.

(Even CrazyEgg’s own dash­board is a one-​​page inter­face, where you can add tests, pages and descrip­tions on the same page, on the fly, all thanks to AJAX.)

One last note:

CrazyEgg is not the only ser­vice of its kind. I also tried Click​Den​sity​.com, which is a com­peti­tor. I pre­fer CrazyEgg, but there are pros and cons to each.

CrazyEgg is more flex­i­ble and usable, but lim­ited to one page per cam­paign. While Click​Den​sity​.com may be a lit­tle clunky, it’s more ver­sa­tile as it can track all pages within a given domain (the tiny snip­pet of code can sim­ply be added to a base tem­plate, for exam­ple, and there­fore added only once).

What I like about Click­Den­sity is that it can track behav­ior through­out a web­site, across many pages, and not just on a sin­gle page. But it’s rather hard to use, and I far more pre­fer CrazyEgg, which is a lot eas­ier to work with.

One tiny snip­pet of code dropped into your HTML (or base tem­plate), and you’re done. How­ever, you must spec­ify each and every page you want to track by cre­at­ing a “new test” for that page.

What I’d love to see is CrazyEgg imple­ment­ing a site-​​wide fea­ture. That would sim­plify things greatly.

But in addi­tion, what I’d like to add to my wish­list is the abil­ity to track traffic-​​to-​​conversions based on click den­sity, lay­outs and, of course, copy (i.e., cal­cu­lat­ing click den­sity ratios by com­par­ing the var­i­ous “hotspots” between land­ing pages to “thank you” pages.)

Granted, I do that right now with some other tools, but it is a tad com­plex and I need to take a few extra steps that I would love to avoid if it were possible.

Want bet­ter AdSense click­throughs? Want higher con­ver­sions? Look­ing for more opt-​​ins? The uses and pos­si­bil­i­ties are too many to list here, and lim­ited only by your imagination.

I cer­tainly rec­om­mend CrazyEgg or any other sim­i­lar tool. You never know, the results might com­pletely sur­prise you — as they did for me.

About the Author

Last 5 Posts By Michel Fortin

• How to Capture and Captivate Attention
• The Real Problem With The Flaw of Attraction
• Two Shining Stars in an Age of Darkness
• Changes, Relaunches, and Extreme Makeovers
• Superior Value Equals Superior Sales

Other Related Posts

• Transparency Versus Hypocrisy
• There's More to Tracking than Conversion
• Why Can't Johnny Get Conversion?
• What Copy Cosmetics Communicate
• Want A Sticky Site? Forget Content!