6 ms·

I think chartbeat on a cool factor is high but on a useful factor it's low and therefore I passed. What web analytics needs is to go beyond just showing the da

I think chartbeat on a cool factor is high but on a useful factor it's low and therefore I passed. What web analytics needs is to go beyond just showing the data (real-time or not) into helping you understand WHAT NEEDS TO CHANGE in order to IMPROVE your website objectives. To be more specific: Tying in a/b testing with user flows and analytics like - ranking pages by abandon rate so you can see where your loosing people - tracking funnel processes like sign up and shopping cart to identify problem areas - ranking links based on clicks to show what users really care about and what is just taking up space on a page - enabling exit interviews and real-time chat with users to collect feedback - and on top of identifying problems it would be really useful for it to present suggestions to help solve the problems. Analytics are just a means to an end but for some reason analytics companies think analytics are the end. Big disconnect here as most of their customers don't have the sophistication to take the analytics and know how to improve their site from them. The analytics company that provides an idiots guide to fixing problem areas and optimizing your site will be a game changer.

arctictony 16y ago
Hey Jon, I'm the GM of chartbeat and totally appreciate your point. Everything you say is super appropriate for the traditional role of static analytics which was to look at usual behaviour on fairly persistent architectures over time and optimize that. However, real-time analytics is almost the antithesis of this. It's focused on exploiting unusual behaviour now often across transient content. With this kind of content and the kind of user flows we are seeing more of across the sites we track (more spikes and troughs than ever before) you can seriously boost your traffic by reflecting user behaviour in the site and (as Jason suggested above) doubling down on what's got the edge in that moment. The page that's driving the majority of your traffic today might not have existed yesterday and might not be looked at tomorrow so purely trying to apply traditional analytics methods to it has diminished utility. So you should totally continue to use traditional analytics for optimising the persistent architecture of your site (I'd highly recommend Kiss Metrics), but think about chartbeat as a complement to that, helping you adapt the transient content on your site to the flows of user behaviour. There's a reason almost every major publisher has integrated chartbeat, some spending thousands of dollars on us a month, and it ain't just because we're pretty ;)
- joncarder 16y ago
  Hey Arctictony, Thanks for the clarification. I agree your guys solution is perfect for media companies or any company that has fresh content coming in and needs to know which content is working. I think Jason's strategy is right on. I guess I'm just bitching because I can't find a "smart analytics" company, one that goes beyond the numbers to recommend solutions for companies like mine. My company is http://www.MojoPages.com http://www.MojoPages.com and we are constantly iterating the site to improve the user experience. Analytics companies like Kiss Metrics don't show the level of detail that we need to really understand the smartest improvement. We're forced to test, test, test and hope we're smart enough to figure out the problem and the optimal solution. Imagine if an analytics company used the data from 1000's of clients to be able to suggest problem areas on your site. An example is if it could analyse your sign up flow and tell you if the location or color of your button or the amount of fields you had etc. was hurting your conversions and the data shows that if you made x,y and z change it would improve 20%. It would make life at a start-up a little easier and whoever creates this magical analytics company a little richer.

arctictony 16y ago

Hey Jon, I'm the GM of chartbeat and totally appreciate your point. Everything you say is super appropriate for the traditional role of static analytics which was to look at usual behaviour on fairly persistent architectures over time and optimize that. However, real-time analytics is almost the antithesis of this. It's focused on exploiting unusual behaviour now often across transient content. With this kind of content and the kind of user flows we are seeing more of across the sites we track (more spikes and troughs than ever before) you can seriously boost your traffic by reflecting user behaviour in the site and (as Jason suggested above) doubling down on what's got the edge in that moment. The page that's driving the majority of your traffic today might not have existed yesterday and might not be looked at tomorrow so purely trying to apply traditional analytics methods to it has diminished utility. So you should totally continue to use traditional analytics for optimising the persistent architecture of your site (I'd highly recommend Kiss Metrics), but think about chartbeat as a complement to that, helping you adapt the transient content on your site to the flows of user behaviour. There's a reason almost every major publisher has integrated chartbeat, some spending thousands of dollars on us a month, and it ain't just because we're pretty ;)

joncarder 16y ago
Hey Arctictony, Thanks for the clarification. I agree your guys solution is perfect for media companies or any company that has fresh content coming in and needs to know which content is working. I think Jason's strategy is right on. I guess I'm just bitching because I can't find a "smart analytics" company, one that goes beyond the numbers to recommend solutions for companies like mine. My company is http://www.MojoPages.com http://www.MojoPages.com and we are constantly iterating the site to improve the user experience. Analytics companies like Kiss Metrics don't show the level of detail that we need to really understand the smartest improvement. We're forced to test, test, test and hope we're smart enough to figure out the problem and the optimal solution. Imagine if an analytics company used the data from 1000's of clients to be able to suggest problem areas on your site. An example is if it could analyse your sign up flow and tell you if the location or color of your button or the amount of fields you had etc. was hurting your conversions and the data shows that if you made x,y and z change it would improve 20%. It would make life at a start-up a little easier and whoever creates this magical analytics company a little richer.