Features
- Record mouse activity on Web pages. This is a tool for webmasters, designers and everyone interested in how users use their Web interface.
- Replay mouse activity in real time. Watch the movements that people have done on your website. This system will replay them either in real time or as a static layer.
- Cross-browser solution. This mouse tracking system is intended to work with every modern Web browser. It was tested on Internet Explorer 6+, Firefox 2+, Opera 8+, Safari 2+, Google Chrome 2+, Camino 1+. Should work also on all webkit-based browsers (e.g: OmniWeb) and mozilla-based browsers (e.g: IceWeasel).
- Not interfere with other scripts. Totally unobtrusive, the system works on pages with any other scripts included.
- Start monitoring when the DOM page is loaded. You do not have to wait for all the images and binary content to be fully loaded for recording the user interaction.
- Work with both static and liquid layouts. The mouse trail is normalized at runtime accordingly to the current viewport and the window size of the previously browsed document.
- Graceful Degradation. Users without JavaScript won't notice anything at all — obviously.
- Walk-Up-And-Use Design. The use of this tool is almost self-explanatory and very intuitive, so that first-time or one-time users can use it effectively without any prior introduction or training.
- Customizable by end users. You can personalize most of the (smt) system, look and feel, etc.
- Check the use of scrolling bars and/or mouse wheel. With (smt) you can check the user's mouse activity in a fine-grained level.
- Check if something has been highlighted, such as a text paragraph. You can visually notice it when replaying user movements, or by analysing the logs from the CMS.
- Generate logs for later postprocessing. A single log file is generated for each visit, so massive postprocessing tasks can be carried out. You can export the user logs in XML files, as well as many other text-based formats, such as CSV.
- Instant logs. Mouse tracking data are stored in the server at the very same time the user leaves the page, so revisions are readily available from that same moment.
- Generate statistics: browser, OS, tracking time and so on. This is intended to be a complement to other system stats, such as Urchin or Google Analytics.
- Filter, mine, refine. Select time series and group database logs by user and browsed page, getting thus the user and page models, respectively.
- Infere users' behaviour. Collected data can be used to find behavioural patterns or even customize a more sophisticated, dynamic Web page.
What's new on version 2?
Here you have summarized the new features.
| feature | (smt)v.1 | (smt)2.0 |
|---|---|---|
| Logs storage | XML | MySQL |
| Cache for HTML logs |  |
 |
| Track DOM elements | |
|
| Visualization API | JavaScript | ActionScript |
| K-means clustering | |
|
| Replay on DOM load | |
|
| Watch all user trails (clickpath) | one per log | |
| Infere page and user models | |
|
| Content Management System | very limited | |
| Multiuser environment | |
|
Ok, but why should i use (smt)?
The great aim is that the system is simple to implement, as much in set up as in use. For that reason, it is not necessary to install additional software neither in the client nor in the server side. It is not either necessary to use a database, although this could be a requirement according to each webmaster needs. !update (smt)2.0 requires a MySQL database, though.
The most direct application is the website usability evaluation from any computer across the world, which supposes a considerable reduction in costs, as opposed to traditional evaluation tests, which usually include in-situ video recording and/or eye tracking. (Both methods need a previous preparation of the environment's work: manual annotations, specific software installation, calibration, timetables, etc.)
Thanks to remote evaluation the user sample is considerably greater, since the number of potential visitors that can enter to a website is huge and more enriching in terms of real data: different users with different computers, different browsers, different computer knowledges... in short a range of possibilities wider than the ones that one can obtain in a lab.
Finally but not less important, the main goal of this application consists of analyzing user behaviour and inferring a visitor profile by means of psychological, social and statistical relations. You can reach this stage when enough representative samples are collected, by looking the user and page models from the CMS.
Finally, this system is open source, free software. You are encouraged to give it a try on your website.
