Discussion on:
View:
Show:
Good concept, but what other options are there for tracking a user through your site? For example, using cookies or javascript, other than tags? More examples? Are there any decent third party add ons that will make this easier and take care of the grunt work?
The author is just giving you a general idea of what you should do to track your site's activity. He is not giving you the actual methods and stuff. For example, if you wanted to do what the article suggests, you would obviously have to store this data somehow, and you would have to choose a method, say for example Cookies, or perhaps you feel you want to have more detailed statistics and create a database to save them. Maybe you want to have them for as long as the site is running, so you store it in the application object. As you can see, this article does not give you the "technique", it gives you the idea. Site Statistics are a bit complex. If you want a third party software, take a look at webtrends reporting tools at: http://www.netiq.com/solutions/analytics/default.asp
My two cents.
My two cents.
WebSideStory (www.websidestory.com) has a third party software HitBox that actually tracks the usage through the site. From the top referers to the exits pages. HitBox uses a small 1 px image on each page to track page views. This process is more accurate than log file tracking. I've been using HitBox for the past 3 1/2 years and I love it.
I don't think any of the above listed applications will tell you what specific link a user clicked on... sure they will tell what paths through your website users take, but they will not tell you whether a user clicked on the main navigation buttonsor the footer navigation buttons, of the same page etc.
That type of data is gold when thinking about redesigning interfaces or websites.
That type of data is gold when thinking about redesigning interfaces or websites.
I think the products like Hitbox are great and we still have clients using the service. The problem is your feeding information into their database and leaving your site, when they are down so are you. We elected to write our own tracking services using a product from cyScape where we can ID the user?s country and browser language to flip to the right website within that language group. With cyScape it gave us all the tools to do everything that Hitbox gave us but now we record it within our client database. As an example only -- visit http://www.jobs4iraq.com/browserInfo.asp where we check out your browser to help us debug problem our client worldwide are having with our site. We are currently having to maintain 12 sites within 11 languages other than English. It is very important to know what your viewers are viewing you with in rebuild your site. Drop my name if you would like more information about the cyScape products.
Contact information:
Sales: 1-800-932-6869
Support: 1-301-424-8200
Contact information:
Sales: 1-800-932-6869
Support: 1-301-424-8200
The basic problem is trying to identify how real people use your system. Start by working with real people.
Capturing statistics is fun and interesting, but it is a substitute to sitting with real people as they try to use the system. Dependingupon the situation, you can go visit actual users at their location, bring users into your location, or use surrogate users. You can interview the users, sit along side them, or indirectly observe them through video cameras or two way mirrors.
The value of the information gathered directly from users exceeds the information that can be inferred from statistics. Go talk to the users first.
Capturing statistics is fun and interesting, but it is a substitute to sitting with real people as they try to use the system. Dependingupon the situation, you can go visit actual users at their location, bring users into your location, or use surrogate users. You can interview the users, sit along side them, or indirectly observe them through video cameras or two way mirrors.
The value of the information gathered directly from users exceeds the information that can be inferred from statistics. Go talk to the users first.
Wayne:
I agree that data gathered from interviews, usability studies, ethnographic research, etc. is an important component; however, it is too time-consuming and expensive to do on an ongoing basis
Site statistics, btw, are generated by real users. Once the site is "tagged", it needs no upkeep and the data can be analyzed in real-time. In addition to being a steady stream of real quantitative data, it is not skewed by the observation process as much of the data gathered during on-site visits or interviews.
So, yes, observing users can be useful; however, it is not very practical considering the time, expense and lack of scale. I find that I get much more actionable information from analyzing "pure" site usage.
sm
I agree that data gathered from interviews, usability studies, ethnographic research, etc. is an important component; however, it is too time-consuming and expensive to do on an ongoing basis
Site statistics, btw, are generated by real users. Once the site is "tagged", it needs no upkeep and the data can be analyzed in real-time. In addition to being a steady stream of real quantitative data, it is not skewed by the observation process as much of the data gathered during on-site visits or interviews.
So, yes, observing users can be useful; however, it is not very practical considering the time, expense and lack of scale. I find that I get much more actionable information from analyzing "pure" site usage.
sm
Yet the problem is that these usage stats show what people did, where they went, when they left, and so on...but they cannot (I say again, they cannot) show WHY people did what they did.
Only observations from people schooled and skilled in the craft can provide valuable data to determine why users do what they do. Does it cost money? Well, as Alan Cooper says, "The only thing more expensive than building software is building BAD software." Without proper user-centered design activities (of which methodological user research is a prime component), software will always meet the wrong target.
So cut $$ from coders and give it to UX developers.
Only observations from people schooled and skilled in the craft can provide valuable data to determine why users do what they do. Does it cost money? Well, as Alan Cooper says, "The only thing more expensive than building software is building BAD software." Without proper user-centered design activities (of which methodological user research is a prime component), software will always meet the wrong target.
So cut $$ from coders and give it to UX developers.
You're right. This data will only give you the "what"; however, you can't determine the "why" until you get the "what."
This approach isn't meant to represent all of the data that needs to be examined when building a site or app. It merely demonstrates how to quickly and easily begin to collect some more useful data.
This approach isn't meant to represent all of the data that needs to be examined when building a site or app. It merely demonstrates how to quickly and easily begin to collect some more useful data.
- Keyboard Shortcuts:
- Prev
- Next
- Toggle
