Comments for Vorpal Analytics http://vorpal.iwright.org/blog Watch where you point those numbers, they could take off someone's head Tue, 20 Oct 2009 16:24:51 -0600 http://wordpress.org/?v=2.8.4 hourly 1 Comment on X Change 2009: Universal Tag Huddle Summary by Fiann O'Hagan http://vorpal.iwright.org/blog/2009/10/01/x-change-2009-universal-tag-huddle-summary/comment-page-1/#comment-8 Fiann O'Hagan Tue, 20 Oct 2009 16:24:51 +0000 http://vorpal.iwright.org/blog/?p=6#comment-8 Matthew, this is a very interesting article, and thank you for mentioning jsHub. As one of the developers on the project, I wish I had been at the Huddle, it sounds like a great conversation. I think you're absolutely right to point out universal data as being one of the foundations of a true universal tag. I have recently blogged about this at https://jshub.org/blog/2009/10/17/universal_tag_or_universal_data/ I agree with VaBeachKevin's comment that it is a fruitless exercise to try and find a mapping from one product feature to a different product. But that is not asking the right question. The real question should be instead, can we identify core concepts such as a product SKU and a selling price, which can then be mapped to an Omniture eVar and to a field in Coremetrics. We should really be looking for these concepts to populate any dictionary, rather than looking at particular encodings of these concepts defined by particular vendors. Matthew, this is a very interesting article, and thank you for mentioning jsHub. As one of the developers on the project, I wish I had been at the Huddle, it sounds like a great conversation.

I think you’re absolutely right to point out universal data as being one of the foundations of a true universal tag. I have recently blogged about this at
https://jshub.org/blog/2009/10/17/universal_tag_or_universal_data/

I agree with VaBeachKevin’s comment that it is a fruitless exercise to try and find a mapping from one product feature to a different product. But that is not asking the right question. The real question should be instead, can we identify core concepts such as a product SKU and a selling price, which can then be mapped to an Omniture eVar and to a field in Coremetrics. We should really be looking for these concepts to populate any dictionary, rather than looking at particular encodings of these concepts defined by particular vendors.

]]> Comment on X Change 2009: Universal Tag Huddle Summary by John Graham-Cumming http://vorpal.iwright.org/blog/2009/10/01/x-change-2009-universal-tag-huddle-summary/comment-page-1/#comment-7 John Graham-Cumming Mon, 12 Oct 2009 10:33:38 +0000 http://vorpal.iwright.org/blog/?p=6#comment-7 There's an enormous difference between Tealium and jsHub. You mention them in the same sentence, but one (Tealium) is a commercial solution, whereas jsHub is an open-source project. In this area open-source is important because it prevents any vendor lock-in, and if you really want the tag to be universal then it has to be vendor-neutral and not vendor-controlled. Additionally, jsHub is proposing public domain standards for web site mark-up so that data is marked up once and read by a single universal tag. There’s an enormous difference between Tealium and jsHub. You mention them in the same sentence, but one (Tealium) is a commercial solution, whereas jsHub is an open-source project.

In this area open-source is important because it prevents any vendor lock-in, and if you really want the tag to be universal then it has to be vendor-neutral and not vendor-controlled.

Additionally, jsHub is proposing public domain standards for web site mark-up so that data is marked up once and read by a single universal tag.

]]> Comment on Communicating Complex Analysis with Google Wave by Rob Ludlow http://vorpal.iwright.org/blog/2009/10/08/communicating-complex-analysis-with-google-wave/comment-page-1/#comment-6 Rob Ludlow Fri, 09 Oct 2009 17:06:34 +0000 http://vorpal.iwright.org/blog/?p=24#comment-6 Great post Matt! Your comment,"...the increased conversion rate is actually artificial due to a decrease in corresponding site traffic" speaks volumes to the complexities of analytics and how easy it is to misinterpret statistics as a whole. I'm so glad I took some of the more advanced stat classes in college because even though I don't remember a lot of the specifics I do remember how easy it is to confuse or misdirect using statistics... for example, it is amazing how often people (even those who we label as experts) confuse correlation with causation. Regarding Wave: I really hope it takes off, if only to really start pushing the envelope of what it means to communicate online. In email and on my forums there is entirely too much room for miscommunication and misinterpretation. I've read that up to 30% of a conversation is lost when body language is unavailable (mannerisms, facial expressions, etc.) and possibly another 40% can be lost when you loose verbal queues like tonality, etc. Hopefully Google Wave can help resolve some of these issues. Keep us posted with your experiences using the platform! Great post Matt! Your comment,”…the increased conversion rate is actually artificial due to a decrease in corresponding site traffic” speaks volumes to the complexities of analytics and how easy it is to misinterpret statistics as a whole. I’m so glad I took some of the more advanced stat classes in college because even though I don’t remember a lot of the specifics I do remember how easy it is to confuse or misdirect using statistics… for example, it is amazing how often people (even those who we label as experts) confuse correlation with causation.

Regarding Wave: I really hope it takes off, if only to really start pushing the envelope of what it means to communicate online. In email and on my forums there is entirely too much room for miscommunication and misinterpretation. I’ve read that up to 30% of a conversation is lost when body language is unavailable (mannerisms, facial expressions, etc.) and possibly another 40% can be lost when you loose verbal queues like tonality, etc. Hopefully Google Wave can help resolve some of these issues.

Keep us posted with your experiences using the platform!

]]> Comment on X Change 2009: Universal Tag Huddle Summary by VaBeachKevin http://vorpal.iwright.org/blog/2009/10/01/x-change-2009-universal-tag-huddle-summary/comment-page-1/#comment-5 VaBeachKevin Thu, 01 Oct 2009 20:01:53 +0000 http://vorpal.iwright.org/blog/?p=6#comment-5 One big problem I see is the variances in the ability of the tools. What I mean is looking at the example of "Coremetrics tagging & Omniture tagging are like French and Spanish", I would say that both French and Spanish have a translation of the word 'book', but Coremetrics may not have something that's comparable to an Omniture eVar. Overall I think it's a good idea but I don't think you could get the big companies to go for it. The idea of it now being so simple to just switch from Omniture to Coremetrics by just changing a couple variables and not having to recode everything on the site will make it much more tempting for a not-100%-happy-customer to do so when their analytics agreement is over. This is something that the larger providers obviously don't want you to do, and I assume would not want to make it easy for you to do so. I do agree with Michael Scherotter's outlook. I think tagging with JavaScript will eventually take a back seat to having the analytics code embedded directly into the page code on the server side. One big problem I see is the variances in the ability of the tools. What I mean is looking at the example of “Coremetrics tagging & Omniture tagging are like French and Spanish”, I would say that both French and Spanish have a translation of the word ‘book’, but Coremetrics may not have something that’s comparable to an Omniture eVar. Overall I think it’s a good idea but I don’t think you could get the big companies to go for it. The idea of it now being so simple to just switch from Omniture to Coremetrics by just changing a couple variables and not having to recode everything on the site will make it much more tempting for a not-100%-happy-customer to do so when their analytics agreement is over. This is something that the larger providers obviously don’t want you to do, and I assume would not want to make it easy for you to do so.

I do agree with Michael Scherotter’s outlook. I think tagging with JavaScript will eventually take a back seat to having the analytics code embedded directly into the page code on the server side.

]]>