Looking forward to your forthcoming book!

Best,
Judah

Thanks again,

Avinash.

I’m glad you got my humor. Thank you for the personal email, and I’m glad to exchange knowledge with you in a reply on your blog. And may I reiterate, imho: your aggregate post was a wonderful contribution of very valuable, hard-learned knowledge. :)

1) As for ClickTracks and VS and my opinion on the vendors. They are all wonderful in their unique ways. And some love NDA’s! But, my opinion, is that one other vendor you named can do what you describe if I read you correctly (i.e. temporal analytics using lookups, bridges, joins). The challenge is how we define “segmentation,” and I’ve been on the ‘net too long to light the flames on that one. Let’s defer to Kotler. But over tea sometime, let’s chat.

2) Re: RFP. I understand and tend to agree with your clarification regarding the perspective of non-intelligence and intelligence. Very well said, but there are many types of intelligence(s) one needs to consider in WA. I knew years ago some programmers who could malloc() in their sleep but tended to look for corners in round rooms (a band once said “going out to find blind spots…”) Your somewhat disparaging past comments about RFP’s were somewhat disconcerting to me, which caused this long time lurker to post. I will agree that RFP’s can be bloated and filled with misinformed questions and could obfuscate understanding of functionality or technology. They also wonderfully irritate vendors. :)

Here’s my quick take (it’s your blog! :) and NOT specific to the web analytics industry or any vendors: RFP’s, like SLA’s, are useful artifacts from the annals of software development. Besides the data and “answers” you receive in responses, the savvy decision-maker looks beyond the data reported by the “yes-only salespeople” and analyzes the response from an anthropological and behavioral perspective. *Directly observed* past and current performance is indeed a predictor of future behavior. The RFP, to me, enables you to observe vendor behavior to answer two important questions to any technology selection process: 1) Can they do the job? and 2) Can I work with them? Anything learned beyond that (and RFP’s are informative when done right) is gravy. Your judgment on the appropriateness and applicability of the response to your business requirements and the data collected sets the stage for the deep technical dives, and my favorite dives–the ones about the usage of statistical methods in the WA tools. And it is great to use the RFP as a documented basis for explaining to internal folks how vendors differ–in the vendors own “vendorspeak.”

If one crafts an RFP question set that meets vendor expectations but with a crafty strategy behind it (i.e not really being so, but appearing to be another uninformed person asking misguided questions from a perspective of “non-intelligence”), you start the relationship “one leg up” because you are underestimated. Thus, you are able to gauge how your future potential partner will treat someone who they think isn’t quite as savvy (like entry-level WA “hires” in the future). And I say “partner” because that’s what you need in Web Analytics. One who returns your phone calls and will work with the personalities of you and your team.

So I say play up that angle if you have the power in your organization to do so; be super, unrelentingly nice; try to come off as naive and ask very simple questions and amazingly complex ones; use the rfp opportunity to ask the same questions in different ways to which you know the answers and so on… to see simply observe vendor behavior. I know that may sound a bit odd, but it works again and again and again across industries, time, economic cycles (if you have the money to spend, authority/time to do so) Look for boilerplate responses, which indicates a lack of careful thought. Which vendor actually took the time to read your response and cared enough to delegate a real human to craft a real response? Perhaps a quality team like VS would even have the Great Peterson guide the response. The side-benefit to all this anthropological/behavioral analysis is that you get real data to analyze about almost everything you want to know and a quasi-“relationship” to proctor into the vendor’s organization to ask deeper questions (on perhaps omitted or glanced over answers). The benefit of the RFP done well, imho, is not only the selection of the best technology match for your requirement. For big companies who will spend lots with the vendor over the long run, the RFP enables the savvy manager to select the best partner to match the organizational culture.

I love Google and have been using it since my days working in what used to be called “information retrieval” when I replaced Northernlight.com with it (back when it did results clustering in the “daze” of talk like “NASDAQ 10000.”) I wholeheartedly agree that if you don’t know anything about WA, then GA is a great place to start before you craft the real RFP you recommend to do the deeper analysis we both require to do our jobs.

ps: I do think GOOG has a very smart cookie that behaves uniquely.

Warm regards and thanks for engaging me,
Judah

Only ClickTracks and VS? C’mon!

(I am open to be proven wrong about this, if there is someone who would like help me with that please email me, blog at kaushik dot net and I will be happy to meet with you or have a phone / email discussion.)

It is my understanding, from my humble experience, that these are the only two tools that will do “post-facto segmentation”. I.E. you don’t have to know all your segmentation possibilities upfront and remember to pass the information to the javascript tag (log file) to be able to answer those questions after data has been collected.

Most tools will need to be told up front and as you know it is very rare that you would know all the questions you’ll want to ask after you get the data. Both of these tools allow you to go back and segment on either Customer Type or Customer Behavior and do so on the fly and fast with lots of possibilities (regular expressions, joins, intersections etc).

The second, and equally important, fact is that you can go back and traverse history and do all of the above for any time period you want. So you can manipulate history and learn and optimize your decisions.

As you evaluate your vendors deeply stress test both of these things to know if they can truly do this. The pat answer will be “of course we can”, then it will be your job to understand what that “we can” really means (you can call me if you want help!).

I do point out in the above post that both of these tools might not be right for everyone, they come with their own unique pro’s and con’s and they each differ from the other in significant ways. End of caveat. :)

That’s like saying deploy Google Analytics at a billion dollar company and don’t do an RFP!

Ok that’s a fair shot. Stated that way! :)

Here is my context:

If you are going to choose a web analytics tool you should do so from a place of intelligence and not non-intelligence. For the most part you will remain non-intelligent about all the massive challenges (that you in your unique company face) unless you actually have web analytics. Actually having a tool makes you smart about a lot of things (tools, processes, problematic IT, crappy sites, problematic Marketers, missing Analysis brain power etc etc).

When you do a RFP you are doing it from a non-intelligence. So why not get a free tool, implement it in ten minutes on your website (don’t use Google Analytics, that’s ok, there are other choices) and start becoming intelligent.

Then do a RFP. You will do a smarter RFP, ask smarter questions, have something (GA or alternative) on the ground on your website to compare your new tools to, and you will make a smarter (and I promise much cheaper) choice.

Does this help? Am I explaining myself?

I know you are only kidding but you asked two really good questions. I welcome your argument and I welcome disagreement and I thank you for caring enough to argue with me. :) Thanks so much.

-Avinash.

That’s like saying deploy Google Analytics at a billion dollar company and don’t do an RFP!

Kidding!

Awesome post, arguable… What good thoughts aren’t?

Judah

“Do you want Reporting or Analysis” can be a full post all by itself. I am sure there will be a lot of debate about that. Thanks for forcing the conversation.