I am a huge proponent of understand the Why (you'll notice that in the book as well). Here are some of the posts where I get on my soapbox…

• Overview & Importance of Qualitative Metrics
• Trinity: A Mindset & Strategic Approach
• Got Surveys? Recommendations from the Trenches
• Lab Usability Testing: What, Why, How Much.
• Solving For The What & The Why: On-demand Webinar
• Multiplicity: Succeed Awesomely At Web Analytics 2.0!

I could keep going on, but I don't want to create a irritating list of links! : )

I am glad you enjoyed your visit to the Chicago Office, I can't think of anyone better to meet there than the absolutely wonderful Ms. Rahn Nomura.

[The other person I think you met must have been the charming Elliott Rader.]

-Avinash.

I had the pleasure of visiting Google last night presented in Chicago by Kristen Rahn Nomura and her co-part [cannot recall name] on Google Analytics She offered a brief but insightful detail to the way the program functions. As the industry continues to evolve the question that has risen is the ability to integrate additional web data-reporting capabilities to ensure [in a nut shell] your site is working properly and you're generating revenue. Often data presented by web analytics lacks Voc [Voice of the Customer] such as Opinion Lab, which complements not only G.A. but also many other Google programs such as Dynamic Comment Cards on Google Ads, etc. If Web Analytics answer the questions Where? and Who? Opinion Lab answers the question: Why?

What are your thoughts?

P.S. I had ordered your book a couple of days ago, and I am very intrigued to learn more.

P.S.S. Your Chicago Google location is awesome. From look and feel, to the people our team acquainted with, it was a real pleasure.

Respectfully,
Mario Bilotas, esm

kind regards,

Brian Tucker, MBA
Competitive Intelligence Analyst

Os recomiendo el siguiente artículo de Avinash Kaushik sobre la relativa importancia de los datos históricos de clickstream.

2) Factores contextuales

Por otro lado, están los factores contextuales. Cuando realicé mi Master en Sistemas de Información en la London School of Economics, una de los mantras que más me repitieron fue “lo importante es situar el sistema de información en su contexto: nada es absoluto, todo es contextual”.
[...]

"Those who ignore history are doomed to repeat it."

By throwing away historical data, I wonder if it increases the risk that people will continue to make the same mistakes again and again on their sites. Quite often there are important lessons that have been learned from this data and while they may lose some relevance, surely they do not become completely irrelevant?

Decay is a fact, but aren't the lessons that are learned from historical data still potentially important to help shape decisions now?

As always, great post which makes me think!
Well, as I can relate to most of the statements (and the comments too- especially Kristen Nomura), I think that the debate is about numbers in general rather than just the historical aspect.

John Marshall tought me how to differentiate between statistics and analytics- -I think this is mostly the point here.

If we simply look at the numbers over the years, we are looking at the statistics, we can find trends – how we grown? have we improved vs. last quarter? (quite similar to financial and SEC reports as stated above) – well these are the flat statistics that our end user usually requests.

I think our job, as web analysts and leaders of this field (well, at least striving to be leaders :-) ) is to teach the user the right questions to ask.

I too meet managers of all levels asking for various numbers and figuers. I think our job is to show them the numbers which really matter. To show them before and after numbers for important events, to show what we CAN do with analytics data.

With regards to changes of tags, systems etc. – well the main problem when we see a change is to know WHY it happened. With so many changes around – design, technology, season, day part, price, customer preferences – it's sometimes hard to tell.

But at this point I'd like to offer a "word" which always helps me to find the good juicy stuff. The word is: "SIMPLIFY".

When we simplify the way we think, when we simplify the numbers and KPI's we will usually get insights. Yes, sometimes we wil not be right. Yes, most of the time we will "miss" very "important" details on the way – but when we simplify we CAN get insights.

So, while I can accept the somewhat frustrating feeling of "too much information" and "too many changes to understand what happened" – But I too think that in our new cyber world one very important rule is to follow what's working in the "oldfashion" offline world – look back, compare – deduct…

(Sorry for this long comment:-) )