Visual Science’s Superior Web Analytics Product

Posted by Marshall on September 25, 2006 | Link It

I spoke with Mark Louck  from Visual Sciences  while I was at the MIXX Conference this afternoon; I viewed the Visual Sciences live Demo and got to ask Mark some very detailed and pointed questions about the Visual Sciences product and decided …..well, read on  and you’ll see what I decided about Visual Sciences.  Just follow my reasoning all the way though this post…it will be worth it.

Visual Sciences, in my opinion, is the best analytics package for a large corporation to run - no question of it.  The point of differentiation (as I pointed out to Mark) is something he told me: "with Visual Sciences you don’t have to tag any pages on the site - a Visual Sciences component will auto discover and categorize (by directory structure) every single page on the site for you without lifting a finger".  That’s a big statement - I accepted it as true as I have no reason not to.

If your running Web Analytics at a large corporation with a couple of million pages - you’d have to touch every single page of the site and put a Javascript tag on it for the analytics platform to pick up the the traffic activity on that page - except with Visual Sciences you would not need to (except for pages where you want to pick up the contents of a form that is being filled out - the main exception).  

Duh!!!!   Do you know how much money that would save ANY large site that’s moving to a new web analytics package? 

Millions of dollars - probably more than the 1 million dollars it costs to set up the Visual Sciences package for a large corporation in the first place (with just 15% per year in the years following the first one - and not including the cost of setting up your servers to run Visual Sciences).   There’s so many sign offs that are normally required in almost any large site - even with excellent page template distribution and in sync CMS systems - touching every page of a web site that has several million pages is one expensive proposition - one that people in charge get panic attacks just thinking about! 

It reminded me of when my old refrigerator died a while back - it was sucking up power and my electric bills were higher than they should have been.  I broke down and bought a new refrigerator and my electric bill went down by half and within a year I had my refrigerator paid for by the cost savings on my electric bill.  

The case of Visual Sciences makes a nice analogy to my refrigerator situation - you buy a much superior web analytics product that saves you more money than you’d have to spent to update all the pages of your large enterprise site.  In essence - for a large enterprise site - it can end up that Visual Sciences is actually FREE ……..that you will have more money in your pocket after one year than if you had spent the money on a competing web analytic product like Omniture or Coremetrics. 

Ok, so much for the cost of deploying Visual Sciences - what about the products analytics and web segmentation capabilities?

Visual Sciences has on the fly URL categorization that any analyst can use to organize parts of a site in a way that makes sense to the analyst and stakeholders and it can be done in real time.   Not only that - but any number of analysts can operate on the same database and make their own URL categorizations that don’t conflict with each other!!! cIt’s clear that Coremetrics can not even approach this level of processing power, nor does Omniture - though Omniture is a close second to Visual Sciences, based on what I have heard.

I wished for this on the fly, real time personal categorization, last year, after running into so many problems running advanced reports - but it’s clear that no other analytics product on the market today can touch Visual Sciences power to categorize data - the very essence of a high end analytics package.

On the the other hand, Visual Sciences is not a web analytics package most companies can afford to buy.  Mark Loucks had only a single page brochure to hand me - as i don’t have a lot of vocabulary to describe what I saw.  Just trust me - I pull data for large sites and I know very well the strengths and limitations of what is out on the market today…..Visual Sciences is the best Web Analytics platform money can buy - no question of it.

In fact, I asked Mark a question:  If Visual Sciences is so great..why don’t you have more clients that run it? 

Turns out that Mark had a reasonable answer;  Until early this year, when Visual Sciences was acquired by HBX - WebSideStory, the company had only one sales person.  After the acquisition - Visual Sciences now has 200 sales persons.

Yet many companies might still think hard on this purchase.  Let me assure you, based on what I saw today….Visual Sciences is the platform to buy - and, in my opinion, it’s the best fit for most corporations to run.  

What makes it possible for Visual Sciences to accomplish this lightening fast processing in real time?   A proprietary database back-end coupled with high end hardware clustering technology.  Most reports have only a couple of seconds delay for any measurement.  There’s no need for programs like "real time monitor" - all Visual Sciences data IS REAL TIME!!!!!!   Also, you can add other sets of data as a batch report (IE: matching up customers by IP Address to visits and purchases on a site - this can be done as a batch job overnight - or whenever you want to run it - asynchronously).

I think I have said enough - but there was much more I saw.  The whole point of Visual Sciences is that it makes collecting and processing data for high end reporting much easier, intuitive and visual than any other package on the market today (guess that is why it is called "Visual Sciences").  With Visual Sciences - you don’t need to spend much time collecting data and can now focus of what your stakeholders really want - ANALYSIS of the data.

Leave it that if you have worked on the kind of web analytics problems and sites I have - reading this will make perfect sense.

 



9 Responses

These are the current comments for "Visual Science’s Superior Web Analytics Product"

Ray McGill
09/28/06 @ 11:54 am

Wow.
Marshall, you really are going out on a limb here. That must of been one hell of a one page glossy brochure.

Based on features promised, I love it too! My IT experience says until I see it implemented on an enterprise level site, it is vaporware. I don’t believe it exists beyond that brochure.

Dynamic - about time
Scalable - doubtful due to processing needs. Despite our advances in hardware, I have yet to see something that can dynamically analyze an enterprise level site. Some tools parse logs into their own consise format and write summaries to parse. But nothing close what I see offered here.

Marshall, you have a great point here. Big Picture = Not touching 1M+ pages pays for tool. If ANY of the other features work as promised, any company is WAY ahead!!!. Ongoing page maintenance and complexity cost $$$.

Visual Sciences, you have a great sales opportunity here! Imagine. Set up your cluster, wheel it into a company, come back the next day and show them what they are buying! Real time! Live! on actual data with no invasive expensive setup. You will own Metrics in a few years.

Anyone see this vaporware work?



09/29/06 @ 11:28 am

I just wanted to clarify that the demo I gave Marshall was on real live customer data not some tweaked out fake vendor data. To your point about scalability and enterprise sites we are successfully providing real time streaming analysis on all metrics and dimensions on several very large enterprise installations today. This is proven technology not something that is yet to be proven out in the real world. On the flip side as an analyst and buyer of tools prior to coming on board with Visual Sciences I can appreciate the seeing is believing mentality. I saw and then I went to work for them. Hope you find this information helpful.



Anonymous
10/01/06 @ 8:33 pm

Marshall,

I hate to be the bearer of bad news, but you’re getting snowballed here. Here are just a few items you should be aware of:

1) Ask the average Visual Sciences customer about time to implementation - you’ll find it takes *many* months, it isn’t instanteous and automagical

2) As for auto-discovery of pages, ask about HTTPS pages and dynamic query-based pages…you’ll find VS doesn’t handle those so well

3) Real-time data? Hardly. VS is based on a sampled methodology. Again, ask existing VS customers about their data and THEY will tell you that their data changes in the interface as they sit there and wait for it to load from the server. Yes, the data actually changes. Try explaining that to your CEO.

4) Categorization of data - You are incorrect here as well. In fact, Visual Sciences customers will also tell you that a major weakness of VS is how limited it is in categorizing data.

5) 200 Websidestory sales people - that is an incredible statement. You should listen to their last quarterly earnings call where I think you’ll hear something quite different.

I could go on and on, there are so many glaring inaccuracies in the the story you were told by Mr. Loucks…I’m sorry you were taken for such a ride. But don’t take my word for it, talk to actual users in the field (not their reference accounts) and you’ll understand what I’m saying.

- Ex-Visual Sciences user and former drinker of the kool-aid



Jakob
10/03/06 @ 9:38 am

Does anyone have any hands on experience with this “wonder tool”?



Paul T
10/03/06 @ 3:08 pm

I’ve used it for nearly two years. Yes it probably takes longer than a month to really get the most of it. But you get the basics you expect immediately. Getting started was less painful than a lot of my experiences with off the shelf software.

As an analyst, I am enthralled with it. It’s simply the most flexible tool I have ever used for any type of analysis. It does sample, but generally we run queries against the full dataset and they complete much faster than any olap tool I have used or sql.

I think perhaps Anonymous above, used an earlier version. Segmenting my data takes almost no effort.



Kool Aid Kid
10/11/06 @ 4:48 pm

WebSideStory barely has 300 employees, there is no chance in hell that 200 of them are salespeople. moreover, VS is a far more complex tool than the products sold by WSS prior to the purchase. The ~25-30 salespeople at WSS are not trained to sell VS products and do not have the authority to sell VS products. they pass leads to VS which is operating as a separate business unit for the forseeable future.

“VS is based on a sampled methodology.” this is incorrect. VS returns data to the user interface as it processes it. so, yes, the numbers change while you watch them. in conjunction with the changing numbers is a confidence meter that indicates how much of the ENTIRE DATASET has been processed to produce the numbers that are being displayed. when the processing is done, the confidence meter reads 100%.



Anon
09/07/07 @ 2:40 pm

Can anyone suggest a great, current web analytics handbook or guide to purchase? There are so many out there, some very out-of-date, and I want to learn as much as I can before I get trained on VS in a couple weeks (I don’t think there is a VS handbook, is there?)
Thanks for advising,



anirban nandi
09/25/07 @ 6:18 am

i am working in a MNC. i am using visual science. cud u plz tell me what is the backend of VS, or what are the flat files it is reading..again from use of VS i understood that VS gives u a summary data. But analytics is playing around with the raw data. so knowledge about the raw data will be appreciated
thank u



11/12/07 @ 8:22 am

Awhile back I spouted off about not liking Avinash’s strategy for ranking “web analytics bloggers” and said something less than complimentary about Marshall Sponder, the WebMetricsGuru blogger. Now Marshall goes and writes one of the most positively effusive pieces about Visual Sciences I’ve ever read titled Visual Science’s Superior Web Analytics Product.



Post a Response

Name (required)

Email (required, not published)

Website (optional)

Note: The following tags are approved for comments on this blog:
<a href=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <code> <em> <i> <del> <strong>