A few months ago, I read a post about a service called Crazy Egg on Tech Crunch. Crazy Egg tracks visitor clicks and creates a great heatmap to show where click activity is happening. What I specifically like about this is that I can set it up to test with, say, 5,000 visitors, then make some changes in response to the feedback, then set it up again to test what effect my changes had. Often with other programs, you can spot problems or areas that need improvement, but it’s difficult to actually see how your changes affected user experience.
Here’s how it works (and why it’s different than Google Analytics), according to the developer:
With Crazy Egg, we are tracking the exact x,y coordinate of a click as well as form fields, submit buttons and other javascript on-click events. Google’s Site Overlay doesn’t track these things. We are also trying to display analytics data visually, through our overlay and heatmap to make it easier for everyone to learn and react to where their visitors are clicking.
I put it on my list of things to try, but I didn’t get around to that until yesterday. Boy, do I wish I did it sooner…
Things I like:
- There is a free version that gives you ample visits (5,000) to play with to see if it’s right for you.
- The plans are pretty cheap - we are doing the $19 per month plan with 25,000 visits tracked. (You do not need to track every visitor. After a certain number of responses, you will notice that the heatmap does not change substantially.)
- You can block specific IPs so the results are cleaner. It will also email people you would like blocked and give them a link to click to opt-out their IP.
- I love that there is just enough information presented, but not too much that it’s hard to make sense of it. It’s very possible to get too much information in stats programs and then not use any of it properly.
Things I don’t like:
- The reporting can be painfully slow to load. It hasn’t affected out load times, but when I go to view the results, I get frustrated at the speed of their website. (And at times it is down for maintenance.)
- Our drop down menus were NOT tracked - which is frustrating because that should be a main place to leave the homepage, but I don’t know how many people actually use it. (More on that below.)
Here are some things to keep in mind when you are interpreting your heatmap:
- Don’t be sad if most of your page is blank. This records CLICKS and not what people are actually reading. If only it could track time before clicks and user sessions, then I would be happy.
- This is NOT the only tool you should be using to try and improve user experience and maximize revenue - this is only thing in your arsenal. Make sure to compare results to your tracking software (Analytics offers some fairly in-depth results on lots of different variables.). Also consider user feedback (which is easy to get on a forum). And finally, if you are really interested, you could always perform some sort of eye tracking testing. (You have to be really hardcore for that one!)
- When you look at the hottest areas, keep in mind that these areas are exactly what your user most wants on your site. They actively pursued what was on the other side of that link. In the world of two second attention spans and millions of sites competing for our attention, *actually* clicking something is a huge deal and should be taken seriously!
- While you do know who is clicking what, you DON’T know why other stuff isn’t being clicked. You will have to use your best judgment on those links. If something isn’t being clicked that you feel is important, it’s time to make some changes and test again. You WILL be able to see if those changes improve the click-throughs on that link once you test again.
Here are the results from our first test at our laptop forum:

As you can see in the upper right corner, just below the logo, there are no clicks on the “Laptop Discussions” menu (only the Discussons and Photos links have menus in them). That drop down menu takes you straight to the forums and is probably the most-used area on the page (as confirmed by Google Analytics’ site overlay). Because no click is required to open the menu, no clicks are registered. This is a major letdown, but there is more to see on the page, so I press on…
My best observations:
- First, there are lots of improvements that can be made based on these results. We will make changes in the coming week or two and then test again. I will post the results when we do.
- I am surprised at how many visitors go straight to the forum search box located just below the banner. That was added after we launched the new design, and it’s the most-clicked item besides the log in boxes. Clearly, searching the forums right when they arrive is important to our visitors. I think that we should make sure that this forum search box is obvious - perhaps moving it to a more prominent position. How many visitors *don’t* see that box and leave when they don’t quickly find what they are looking for? The only way to know that is to tweak and then test again.
- In the same vein, I know that our search is not that great. It’s a weakness of vBulletin. Once you hit a certain post count (we are over 2,000,000 now), the search just doesn’t work that well. It concerns me to think that so many visitors try to search and probably don’t get the best results. This has worried me for some time, but this reminds me that we need to make sure there isn’t a way to improve the searching experience so we don’t loose visitors with short attention spans and the inability to dig in the forums for information.
- We added the three boxes at the top of the page with this redesign. They were put there to create ‘buckets’ to sort our visitors into. The idea is to satisfy the main needs of our users right when they get to the site. We know that our visitors (those who need navigational help - this doesn’t apply to regulars) are researching, shopping, or just looking for a forum about laptops. From this heatmap, that was clearly a good idea because they are being used. There are no clicks on the Community box, I presume, because the Notebook Discussions menu is used. I can’t be sure of this unless we come up with a way to track that menu. With these results, I know that these boxes are serving a purpose, so the next step is to test and tweak the laptop research and laptop shopping pages on the other side of those boxes. I know that there is a ton of room for improvement on both of these pages, so I look forward to doing a baseline test and then tweaking further. I wish Crazy Egg tracked user session lengths… that would be so perfect.
- It’s worth noting that a large chunk of visitors are already registered and immediately log in. So what does this mean? Do we need more fresh blood? (40% are returning visitors each day - that seems like a perfectly suitable percentage for a forum.) Or does this mean that we are simply losing a lot of people who never click? (And we need more data to come to that conclusion - Analytics would be a good tool as it shows session lengths.)
- The final main observation is that people want to see the notebook reviews. Of course, I knew this already, but the heatmap confirms the importance of this feature to our visitors. We recently rewrote the whole laptop review database to make it more useful and easier to add to, but we haven’t started really promoting it to get reviews. A competitor actually pays $50-$60 per review. That isn’t a playing field I want to be on, so I plan to get creative. Don’t get me wrong, members do submit reviews, but there is a lot of room for improvement (is that a theme here?
). This is actually one thing I would love to tackle at the Elite Retreat next week. There is a marketing angle we are missing here. I would bet we aren’t vocal enough about the fact that we want reviews. I bet we could make it easier to submit them. I bet that we could create a top-notch review database that people really want to be a part of - we just haven’t really gone there yet. Stay tuned on this one. (But the pressure is on seeing this heatmap!)
One thing that I should mention is that there are two other views of these same results on the Crazy Egg Dashboard: a list view that shows actual click numbers, and an overlay view that shows pluses and minuses that expand and contract to show clicks and their percentages. I find the heatmap the easiest to interpret, but I do toggle back and forth for more specific information.
I feel like this post is a bit self-absorbed since it’s all about how to make my forum better, but anything said here applies to you all as well. We *all* could stand to make some improvements, so apply some of these ideas to your site. As you are making adjustments, keep in mind that small changes make big differences, so try not to change a bunch of things at a time. Change one or two things and then test for results. If you don’t get the desired reaction, try again. And remember: your changes could end up creating negative experiences for your users, so carefully watch what you do. If you are really ambitious, you can create two versions of your page and test both concurrently.
I am getting ready to go to the Elite Retreat on Sunday, and this tool will surely help me better prepare me for the conference. I have already identified areas that need critical attention, and this is just the beginning. If you start testing and tweaking with Crazy Egg, please just be very conscious of your users and their needs. Please don’t go nutty changing your whole site around. And when you get results, please share them with us!
Happy Testing and Tweaking!
UPDATE: CrazyEgg has generously offered all ForumTrends readers a free upgrade to 10,000 visits per month plus live reporting. As I said in my post, this is an ample amount of visits to test and tweak your site to make a real impact. To take them up on their offer, simply email info@crazyegg.com and tell them that you are a ForumTrends.com reader and you would like to use their free offer. This is a great deal - I hope it helps you improve your forum!
by Laua Alter, NoteBook Forums
Very nice article Laura. Don’t be overly concerned about the being ’self-absorbed’ bit. I actually prefer seeing concrete examples based on someone’s personal experience. I’m sure that the readership of this blog are quite capable of figuring out how to apply it to our own situations. Thanks very much for taking the time to do this, and from the looks, you invested a significant chunk of time into doing this.
You might also want to look at a service called ClickTale, it actually records the actions of visitors so you can see what paths they take, www.clicktale.com
“The reporting can be painfully slow to load. It hasn’t affected out load times, but when I go to view the results, I get frustrated at the speed of their website. (And at times it is down for maintenance.)”
I’m trying the trial as well and have noticed how VERY VERY slow this service is when it comes to viewing your reports. I’d be interested in purchasing one of their plans if the pages would actually load. :\
Bryce, Thanks for the link. I’m going to check it out myself as well.
-Michael
Hi Guys,
We are working on our speed issues. Our recent downtime was all related to upgrading our servers and speeding things up. Feel free to contact us if you have any further questions or specific issues with your tests, even speed issues (many times they are easy fixes)
it would be a great tool if the free feature went from 5000 to
more like 70,000 because 5k is only a hand full of people compared to how many peope i get to my blog a day, and one day is always different from the next..