acne scar excision…

bisena.com…

is diflucan safe to take while pregnant…

Gez

Dan here (from the people behind clickdensity). Thanks for the interesting comments – I’ll reply to as many as I can (and feel free to throw any more my way too).

First off, cost is always an issue, so I’d better tackle that. We’ve written a bit about this on the site (we try to be as open as we can about it), but I’ll repeat some of it here…

We’ve tried to make clickdensity one of those rare breed of Web 2.0 products that actually has a business plan beyond ‘get traffic, make loss, sell out’. We’re open about how we calculate the pricing model – it’s basically just costing based on amount of server resource taken up. We don’t have the scale (server farm) of Google or the bigger players (yet), but as the service does grow (and the resources behind it do too), the economies of scale should allow us to offer more for less.

If anyone has any interesting/novel ideas for costing models, we’d love to hear those. And, as Kevin Burton says, we are able to offer full-time support because of the current (paid-for) model. We try to reply to support within a few (working) hours, but due to being techcrunch’ed, you may experience a slight delay over the next few days!

As others have pointed out, although the ‘hover maps’ feature is /similar/ to Google Analytics overlay, there are some differences, and the main features (heat maps, click maps) are not at all similar.

For the technologists – we’re using JSON, not AJAX (due to the cross-server nature of communications needed to record the data). Although we did use AJAX when we prototyped the idea locally.

Footprint-wise: yup, about 9Kb. It actually (until fairly recently) used to be about 4kb, but then we had to add a hashing algorithm to better identify unique objects on the page, which doubled the size! Again, any ideas to reduce the size – we’re open to suggestions!

For those concerned with privacy – we of course take this very seriously. We’re implementing P3P today, which goes some way to making our service more open/transparent. We’re not collecting keystrokes or anything of the sort – in fact, pretty much the only thing we’re collecting on top of ‘normal’ server-based stats is X/Y of click and the ‘text size’ setting of the browser. We also don’t provide any means to identify individual users in the reporting interface, and in fact the premise of the product implies that generic patterns are more important anyway.

Yup, Timothy/Josh, we’re pretty sure that the people clicking in ‘whitespace’ are usually just getting focus back into the current browser. Although you do often see clicks in paragraphs of text, where some people follow the text with their mouse when reading (and also at the start and end of text, where people copy/paste text from a website). Although the results are a little out of date, we initially did some analysis on the (prototype’s) data here: http://www.orei...y_coordina.html

Re: Robbin Steif talking about Google’s ‘click by goal’. We also offer a similar function, where you can only plot the data from people who ‘came from’ and/or ‘went to’ a specific URL (e.g. went to ‘pricing’ from ‘benefits’, or people who came from X and went to ‘Sign up’).

I think it’s worth me just pointing out that what you see is very much the first phase of a long-term project for the service. Due to it being a paid-for service, we’re able to commit full-time developers to the project, and we release new functionality every couple of weeks. We have some /major/ new features planned, all of which are complementary to the usability-type analysis that we hope it’s already providing. There are some really exciting things coming up in the next few months…

Don Wilson: yup, it can handle that traffic. I’m not sure how much I can talk about who’s currently using it, but we have one of the largest high-street retailers in the UK, a leading online bank, and many others with million+ page views a day.

It’s worth pointing out that you don’t need to record ‘every’ click in order to get useful results, and you don’t even need clicks over a number of days. As the main purpose is to analyse (and help to continually improve) the usability/IA of websites, it’s not so important that you record a duration of clicks. Rather, it’s more important that you have /enough/ clicks to demonstrate ‘average use’ of your site (i.e. to make useful heat/click maps). So, although the Starter package may only have a 10,000 click limit at the moment (which you can wipe/start recording again at any point), this should be more than enough for most small websites to produce useful heat maps/results.

If anyone has any ideas for new features or improvements, please send them through – we’ve been suprised by how many great suggestions people have for cool new functionality. Which reminds me, one of the things we’re planning is an API and distribued plug-in architecture, which could provide some really interesting results…

Thanks for reading my waffle,

Dan

http://blog.cor...cking-clickmaps
http://blog.cor...-make-heat-maps

If anyone wants some more information or is interested in a more complete solution, drop me a line and I’ll try to write about that.

I didn’t look at the size of the file but for those of you who are worried, you might like this best when you already have server side analytics instead of client side, since your page doesn’t already have “extraneous” javascript on it.

In Google’s defence: When their overlay works, it shows you click by goal. (e.g. 55% of the people who clicked on this link also hit my goal 1: sign up for email marketing) Also, Simon, there is a little bit of setting up to do in GA to be able to see the difference between links to the same page from the same page (e.g. both the picture and the text link to the same page) but not a lot. ROI Revolution did a great piece on their blog. All you need to do is add an additional query parameter to the link. Here, you can read it yourselves:
http://www.roir..._to_figu_1.html

I don’t think that it is invasive, as long as they aren’t tying those clicks to IP addresses. Th

What I am not too sure about is the privacy implications of this. With timings as well sounds a bit roo invasive and almost like key logging, and I wouldn’t want to put that on anyones system whether for a legit purpose or not.

第一次看见 Clickdensity 演示的时候，感觉有点像 MapSurface， MapSurface 当时提供内嵌计数统计面板的时候，有设定页面各链接的点击数和频率。当然了，Clickdensity 远远不止这些简单的功能。……

Anyway.. god bless them for charging for this! (I’m serious). I really need this feature for Tailrank and A/B testing. This way if it breaks I have the right to get angry

Serious… I’ve been meaning on coming back and trying CrazyEgg again.. they had a bug which prevented me from using it in production at the time… (might have been my fault)… so now I have to re-open this can of worms.

Kevin