Implementing and developing a leading web analytics tool

Disclaimer: This post is NOT intended to put large web analytics vendors in bad light and to endorse the use of Google Analytics instead.

In my daily work I rely on one of the larger web analytics solution vendors, which over the past two years has provided me with rich data, custom and on the fly reporting and one the whole things were alright. Until we needed better reporting, which we thought would be relatively easy to get since we chose one of the leaders in the market. Little did we know.

I’ve noticed that over the past couple of years it’s become increasingly difficult to work with these guys. It was already quite a feat to get through implementation, which was almost completely unguided and parts of which had to be corrected later on. Bit by bit we learned more about how the solution works and once we deciphered most of the ingeniously chosen terms used solely by this vendor we started to uncover what went wrong during implementation and we obviously needed to correct these issues to get everything sorted.

Later on, when the time came to renew the annual subscription, we ran into more trouble which made me realise we were pretty much stuck with this vendor because there’d be no way in hell the chief would let us switch to something completely else. The problem, this time, came from the vendor’s sales execs. They were apparently too busy reeling in new customers to be bothered to send us a renewal notice and invoice and we had to remind them to make an easy sale and some good commission I imagine. Extremely arrogant, but well, we’d become dependent on these guys and it was pretty much a case of make do and try to get along.

The real trouble started when we started to realise there were serious flaws in our reports, the main one being conversion funnels based on page views. Originally, key pages on our corporate sites (paths that lead to our conversion goals) were tagged with eVars and events, which enabled us to segment and cross-reference data. It also allowed us to set up funnel reports (visual of conversion paths) and start following changes over time. Great, if it weren’t for the fact that by default the vendor had set these event tags up to be triggered with every single page view, seriously skewing our data. Now in some cases this isn’t too big a deal, but the way our corporate sites are structured and used by visitors this just means unreliable data, clear and simple.

After having paid a certain amount of development hours, at a ridiculously high rate I might add, we were contacted by an affiliated consultant who would guide us through the process of development for the items we needed to have corrected. To cut a long story short, six weeks and three failed attempts later we’re finally testing new javascript code which is giving us the results we need.

My conclusions

1. Advanced functionality comes at a price that can’t be expressed in dollars, euros or yens
2. The more advanced web analytics tools are not always by default “better” tools. You need a lot of kneading and a lot of luck and persistence (getting the right consultation from someone with experience with your situation) to get the reporting that you require
3. Implementation of a leading web analytics tool is a continuous process, as web analytics develops (into business intelligence)
4. Working with these guys means the concept of “deadline” is completely meaningless

My advice to those who are yet to choose a web analytics vendor

During your courtship with vendor X, ask for, nay DEMAND:

• Full disclosure on what implementation will entail (and make sure your tech guy is up to scratch)
• A set amount of development hours included in the set up fee (at least 30, depending on the complexity)
• A fixed price for consultation hours beyond the initial implementation phase

Had any bad experiences with vendors yourself? Leave a comment!